
2019 | 0 views | 0 downloads | 126 Pages | 3.93 MB
blended with selected PEO and PEO-PDMS copolymer additives with improved separation performance and long term performance sustainability; (2) the 2nd generation membrane: composite hollow fibre membranes developed in this study with multi-layer coating using selected CO2-philic PEO-PA block copolymers (PEBAX ) as selective layer.
CO2CRC PARTICIPANTSCore Research Industry Government SupportingParticipants Participants ParticipantsCSIRO ANLEC R D CanSyd AustraliaCurtin University BG Group Charles Darwin UniversityGeoscience Australia BHP Billiton Government of South AustraliaGNS Science BP Developments Australia Lawrence Berkeley National LaboratoryMonash University Brown Coal Innovation Australia Process GroupSimon Fraser University Chevron The Global CCS InstituteUniversity of Adelaide State Government Victoria Dept of State University of QueenslandUniversity of Melbourne Development Business InnovationUniversity of New South Wales INPEXUniversity of Western Australia KIGAMNSW Government Dept Trade InvestmentWestern Australia Dept of Mines and PetroleumEvaluation of CO2 capture with highperformance hollow fibre membranes fromFinal ReportANLEC Project 3 1110 0087Dr Hongyu Li Professor Vicki Chen Jingwei Houand Dr Guangxi DongFebruary 2015CO2CRC Report No RPT14 5254AcknowledgementsThe authors wish to acknowledge financial assistance provided to the CO2CRC by the AustralianGovernment through its CRC program and through Australian National Low Emissions Coal Researchand Development ANLEC R D ANLEC R D is supported by Australian Coal Association LowEmissions Technology Limited and the Australian Government through the Clean Energy InitiativeCO2CRC LimitedSchool of Earth Sciences University of MelbourneLevel 3 253 283 Elgin Street VIC 3010PO Box 1182 Carlton VIC 3053p 61 3 9035 9729www co2crc com auReference Li H Chen V Dong G and Hou J 2015 Evaluation of CO2 Capture with HighPerformance Hollow Fibre Membranes from Flue Gas Final Report ANLEC report CooperativeResearch Centre for Greenhouse Gas Technologies Canberra Australia CO2CRC PublicationNumber RPT14 5254 pp 109CO2CRC 2015Unless otherwise specified the Cooperative Research Centre for Greenhouse Gas TechnologiesCO2CRC retains copyright over this publication through its incorporated entity CO2CRC Ltd Youmust not reproduce distribute publish copy transfer or commercially exploit any informationcontained in this publication that would be an infringement of any copyright patent trademark designor other intellectual property rightRequests and inquiries concerning copyright should be addressed to the Commercial ManagerCO2CRC PO Box 1130 Bentley WA 6102 AUSTRALIA Telephone 61 8 6436 865Executive SummaryThis final technical report is prepared for ANLEC R D Project 3 1110 0087 This projectaimed to fabricate high performance hollow fibre membranes for CO2 capture from flue gasesand to assess their performance with both a laboratory synthesised gas mixture and real fluegases from a power plant In concluding this project we were expected to1 Select one or two polymers and additives commercial polymers as materials forproduction of hollow fibre membranes with potential for superior performance based onthe CO2 permeability and CO2 N2 selectivity2 Develop 2 hollow fibre membranes with improved CO2 permeability of at least 50higher and comparable CO2 N2 selectivity compared to benchmark hollow fibre3 Evaluate the tolerance of the hollow fibre membranes to impurities in flue gas with theobjective of achieving stabilised selectivity and permeance over one month operation4 Test the performance of membranes developed in this project with real flue gas on sitea power plantThe project started with an extensive State of the Art assessment of material selection andbaseline performance criteria that considered the major techno economic issues for largescale deployment Fundamental technology background for membrane gas separation andits application in CO2 capture particularly in post combustion flue gas CO2 capture werereviewed This was followed by identification of benchmark membrane performances basedon materials that are currently being synthesised and fabricated at scale in hollow fibreconfigurations Those materials were poly p phenylene oxide polyimide Matrimid andpolyethersulfone PES They exhibit a permeance in the range of 50 GPU and CO2 N2selectivity of 25 In conjunction with the good mechanical properties and manufacturingmaturity of these materials and in line with the use of MEA as the solvent benchmark for CO2capture these polymers were considered to be the benchmark materials in this study Assuch the CO2 N2 separation target for this project was set as the CO2 permeance surpassing50 GPU and CO2 N2 selectivity over 25Based on the extensive review screening and selecting benchmark polymer materials weselected two materials for hollow fiber membrane development for laboratory and on sitetests with flue gas1 the 1st generation membrane hollow fiber membranes fabricated using Matrimidblended with selected PEO and PEO PDMS copolymer additives with improvedseparation performance and long term performance sustainability2 the 2nd generation membrane composite hollow fibre membranes developed in thisstudy with multi layer coating using selected CO2 philic PEO PA block copolymersPEBAX as selective layerMembranes of both generations were fabricated in house with their separation performancetested with clean CO2 and N2 pure gases no impurities and CO2 N2 gas mixture in thelaboratory For the 1st generation membranes improved CO2 permeance between 24 34GPU and CO2 N2 selectivity between 30 40 than commercially available products wereFor the 2nd generation membranes a new protocol for dissolving Pebax 1074 grade polymerusing simple and environmentally friendly mixed solvent solution was developed followed bysystematic studies on the phase structure of Pebax dense membranes including blendedmembranes and their related gas separation performances Based on this fundamentalknowledge composite hollow fiber membrane development was conducted through selectionof suitable microporous substrates selection of materials for protective gutter layer anddesign and construction of a unique dip coating facility funded by CO2CRC suitable forhollow fiber composite membranes At the best combination of the conditions screened inthis study CO2 permeance up to 560 GPU and CO2 N2 selectivity above 46 was achieved atroom temperature whereas 950 GPU and CO2 N2 selectivity of 30 was achieved at thecommonly reported temperature of 35 C This performance was better than the bestreported results for composite hollow fibers for CO2 captureIn the Phase 2 membrane development the separation performance of the candidateMatrimid hollow fibre membrane was evaluated in the laboratory for tolerance to NOimpurity the primary impurity present in the flue gas after the pre treatment column andwater by testing with a synthesised CO2 N2 NO gas mixture with addition of water vapourThe test results indicated that the trace amounts of NO only had minor impact on the CO2 N2separation performance for the Matrimid hollow fibre membrane with 4 Silwet L 7607Both CO2 permeance and CO2 N2 selectivity dropped less than 10 compared with themixed gas permeation results without NO However the performance tested with humidifiedgas gas feed passing through a water humidifier to add water vapour to the feed to themembrane indicated severe reduction in CO2 N2 selectivity 70 and limited CO2permeance up to 16 at water vapour activity between 0 6 and 0 86 stressing theimportance of water removal pre treatment process in membrane applications for flue gasIn evaluation of the composite hollow fiber membranes we observed that similar to that ofMatrimid based hollow fiber membranes the presence of NO did not affect the membraneseparation performance significantly The presence of a small amount of water at low activityof 0 08 and 0 16 had an insignificant influence on the separation performance while theevaluation at higher water activity level was not conducted due to the restricted resource inthe lab environmentWith the purpose designed and constructed mobile membrane test unit the on site test withthe 1st generation Matrimid hollow fibre membrane was conducted at Delta Electricity atVales Point with untreated flue gas as the expected pre treatment facility linked to the othercapture plant on the same site was unavailable Seven membrane modules were preparedwith 5 modules tested on site A decrease in both CO2 permeance 15 GPU at the highestand CO2 N2 selectivity up to 15 in comparison with the results obtained with pure gases inthe laboratory was observed However 2 of them both with 4 Silwet additive fabricatedwith 15 cm air gap exhibited minimal loss of separation performance after 3 days operationwith untreated real flue gas indicating good integrity against real industrial conditionsDespite the good chemical and mechanical stability of the 1st generation membraneprolonged tests with flue gas on site was discontinued due to the interrupted supply of fluegas caused by power plant maintenanceWhen the flue gas supply was resumed the subsequent on site tests were conducted withthe 2nd generation composite hollow fiber membranes because much better performance hadbeen observed for the second generation membranes in lab tests The on site test of the 2ndgeneration membrane composite hollow fiber membranes made with polyvinylidene fluoridePVDF microporous fiber as substrate coated with multiple Polymer poly 1 trimethylsilyl1 propyne PTMSP as gutter layer and PEBAX as selective layer were conducted withthree membrane modules that had been evaluated in lab tests In the first 14 days of testsminimal pre treatment of the flue gas feed was facilitated through regular change of thedesiccant column and draining of the water trap bottle used for collection of condensedwater in the piping line Relatively stable permeance and selectivity were observed with allthree modules with CO2 permeance of 90 120 GPU and CO2 N2 selectivity of 3 5 While theCO2 permeance and the CO2 N2 selectivity was lower than what was achieved in the lab withsynthetic gas mixture the mechanical integrity of the membrane was maintained through theflue gas exposure in that when the membrane module was brought back to the UNSW anddried followed by testing with pure gas only 12 reduction of CO2 permeance from 500 to441 GPU and 5 reduction in CO2 N2 selectivity 31 2 to 29 6 was experiencedWhen the membrane was subjected to the flue gas without pre treatment severe loss ofpermeance and selectivity of all three modules were observed and permanent damage tothe membrane mechanical integrity was suspected as evidenced by the irreversible reductionof membrane selectivity after drying tested in the lab The damage to the membrane wasmost likely due to flooding of the membrane module by condensed water in the feed lineIn conclusion 2 generations of membrane were developed and tested with both lab and onsite conditions Improved CO2 separation performance was achieved with the first generationmembrane compared with existing membranes while the 2nd generation compositemembrane achieved excellent separation performance with potential to make a membraneprocess competitive for CO2 capture On site test results for both generation membranesdemonstrated that the stable membrane separation performance could be achieved butperformance was severely impacted when subjected to flue gas without pre treatmentFlooding of the membrane module by condensed water in the pipeline could causeirreversible damage to the membrane fibersThese observations suggest that pre treatment of flue gas particularly removal of water isessential prior to feeding to the membrane system Systematic evaluation of the influence ofmembrane performance by water vapour should be conducted through well controlledexperiments In addition the stable performance of the hollow fibre membranes in the fieldindicate that further development of membranes with improved characteristics is highly likelyto lead to a membrane process with suitable separation performance for flue gas treatmentExecutive Summary iList of Tables iiiList of Figures iv1 Introduction 11 1 Description of the project 11 2 Milestones and deliverables 31 3 Layout of this report 42 Membrane Gas Separation for CO2 Capture A state of the art review 62 1 Membrane separation mechanisms 82 2 Membrane structures 92 3 Selection of membrane materials 102 4 Membrane fabrication process 162 4 1 Phase inversion 172 5 Hollow fiber membrane fabrication 182 6 Fabrication of composite membranes 192 6 1 Fabrication of thin film composite TFC membrane 192 7 Evaluation of membrane performance 212 8 Constraints of large scale implementation 222 8 1 Physical ageing 222 8 2 Membrane plasticization 242 9 Effect of minor components 25Water vapor 252 10 Economic considerations 262 11 Membrane process design 272 12 Research benchmarks 303 Development Hollow fiber membranes using Matrid blended with selected additives1st generation membrane 323 1 Hollow fiber membrane fabrication 323 2 Gas permeation tests 353 2 1 Evaluation of PEG additive on gas separation performance 353 2 2 Evaluation of PEG PDMS additive on gas separation performance 373 3 Effect of additive on membrane CO2 plasticization 384 Development of composite membrane for CO2 capture from flue gas 2nd generationmembrane 414 1 Fabrication and evaluation of PEBAX dense film 414 2 The effect of blending on PEBAX dense membrane structure and separationperformance 454 2 1 The effect of pressure and the performance with mixed gas 514 3 Development of thin film composite TFC membranes 534 3 1 Flat sheet TFC membrane 534 3 2 Composite hollow fiber membranes 544 3 3 Screening the substrate 554 3 4 Stability of PTMSP gutter layer 604 3 5 Comparison with TFC membranes in literature 615 Membrane performance in the presence of NO and water vapour 635 1 Matrimid blended hollow fiber membranes 635 1 1 Effect of NO in the feed mixture 635 1 2 The effect of water vapour and temperature 675 2 Pebax composite hollow fiber membranes 706 Pilot Plant Design 746 1 Feed composition property 746 2 Selection of equipment 756 3 Nomenclature 837 On site tests in Vales Point Power Plant Milestone 4 847 1 On site installation 847 2 Floor Plan for the CO2CRC Membrane CO2 Capture Facility 867 3 On Site Test of the 1st Generation Membrane Milestone 5 877 3 1 Raw Flue Gas Composition 877 3 2 On Site Tests of blended Matrimid hollow fiber membranes 897 3 3 Results from On site Test 1st generation membrane 907 3 4 Modifications on the Flue Gas Feed Inlet Connection 937 4 On Site Test of the composite hollow fiber Membranes Milestone 6 947 4 1 Modifications on the membrane unit for the 2nd generation membrane test 947 4 2 Results from on site test 2nd generation membrane 958 Conclusions and recommendations 1018 1 Conclusions 1018 2 Recommendations 1049 References 105List of TablesTable 1 1 Key milestones and specific tasks 3Table 2 1 Membrane materials used in industrial scale gas separation applications 6Table 2 2 CO2 N2 gas separation properties for variety of membrane materials 13Table 2 3 Maturity of membrane development in gas separation applications 15Table 2 4 Packing density of typical membrane module configuration 17Table 2 5 Summary of the polymeric membrane materials for CO2 capture from flue gas 31Table 3 1 Gas permeation test results for Matrimid hollow fibers with 0 4 8 and 12 wt PEO PDMSCopolymer conducted using pure gases 37Table 4 1 CO2 permeability and CO2 N2 selectivity of Pebax 1657 and 1074 membranes cast withdifferent polymer concentration solutions Gas permeation tests conducted at 200 psi and 35 C 43Table 4 2 CO2 permeability and CO2 N2 selectivity of Pebax 1657 and 1074 dense membranes castwith different solvent evaporation rate 200 psi and 35 C 44Table 4 3 Thermal properties of Pebax 1657 blend membranes with IM22 and Silwet 49Table 4 4 Thermal properties of Pebax 1074 blend membranes with IM22 and Silwet 50Table 4 5 CO2 separation performance of PES PEBAX composite membrane 54Table 4 6 Water flux of hollow fibre substrate pump rate 20mL min 55Table 4 7 CO2 permeability and CO2 N2 selectivity of PES and PVDF substrates coated with gutterlayers only tested at room temperature 56Table 4 8 Gas permeation test results of composite hollow fibers with PES and PVDF substrates 57Table 4 9 CO2 permeability and CO2 N2 selectivity of composite hollow fiber membranes usingPTMSP as gutter layer 60Table 5 1 Percentage change in CO2 and CO2 N2 selectivity for humidified pure gas 69Table 6 1 Flue gas composition at the Vales Point Power Station 74Table 6 2 Specifications of the major equipment 76Table 6 3 Potential hazards causes consequences and controls 77Table 6 4 Operating conditions of the membrane capture pilot plant 78Table 6 5 Process conditions and gas composition for each line 79Table 7 1 Flue gas composition at Vales Point Power Station 88Table 7 2 CO2 N2 separation performance of the selected membrane modules 90Table 7 3 Composite hollow fiber membranes developed in this study Selected modules fortests on site with flue gas were highlighted 96Table 7 4 Comparison of the membrane pure gas test results in lab before and after the on site test 99List of FiguresFigure 2 1 Mass transport mechanisms in pressure driven membrane systems 8Figure 2 2 Physical structure and morphologies of common membranes 10Figure 2 3 Robeson upper bound correlation for CO2 N2 8 12Figure 2 4 Methods of membrane fabrication process 16Figure 2 5 SEM image of in house fabricated Matrimid asymmetric hollow fibre membrane 17Figure 2 6 Schematic of a typical hollow fiber spinning process 18Figure 2 7 Structure of multilayer TFC membrane and scheme of dip coating facility for hollow fibreproduction developed during this project 20Figure 2 8 SEM image of a hollow fibre membrane with a coated dense layer formed using a dipcoating technique 21Figure 2 9 O2 flux profile of PES hollow fibre membrane as a function of time 57 23Figure 2 10 CO2 permeation isotherm as a function of feed pressure pure CO2 feed stream 66 24Figure 2 11 Effect of Membrane selectivity on capture cost of CO2 4 27Figure 2 12 Effect of CO2 permeability on the cost of CO2 capture for different vacuum membranesystems single stage membrane system two stage cascade membrane system and twostage cascade membrane system RR 86 28Figure 2 13 Stage gates to large scale implementation 29Figure 3 1 Chemical structures of Matrimid Silwet L 7607 and PEG 400 33Figure 3 2 Laboratory scale hollow fibre spinning equipment located at UNSW 34Figure 3 3 In house fabricated Matrimid hollow fibre membrane and membrane module with 4 to 5strains of fibers potted inside the stainless steel tube 34Figure 3 4 Schematic representation of gas permeation test rig 35Figure 3 5 Gas separation performance for Matrimid hollow fibers with 0 4 8 and 12 wt PEG Thegas permeation tests for CO2 and N2 were conducted at 6 bars at room temperature 36Figure 3 6 CO2 N2 mixed gas 22 78 vol separation performance closed symbols are pure gasresults and open symbols indicate the mixed gas results 36Figure 3 7 CO2 N2 mixed gas 22 78 vol separation performance closed symbols are pure gasresults and open symbols indicate the mixed gas results 38Figure 3 8 CO2 permeance as a function of feed pressure for Matrimid hollow fibers with differentPEG contents The arrows indicate the estimated plasticization pressure 39Figure 3 9 CO2 permeance as a function of feed pressure for Matrimid hollow fibers with differentPEO PDMS copolymer contents the arrows indicate the estimated plasticization pressure 39Figure 3 10 CO2 permeance over time under a constant pressure 20 bar for pure MatrimidMatrimid with 8 wt PEG and Matrimid with 8 wt PEO PDMS copolymer membranes 40Figure 4 1 Chemical structure of general PEBAX PA and PE changes with the grades of particularproducts 41Figure 4 2 Phase images of Pebax 1657 a and 1074 b dense membranes obtained from SPMscanning 43Figure 4 3 CO2 and N2 permeability of Pebax 1074 membrane in gas mixture solid line comparedwith pure gas dash line the initial value the value at 24hours 44Figure 4 4 CO2 N2 selectivity of Pebax 1657 and 1074 membrane in gas mixture Pebax 1657Pebax 1074 45Figure 4 5 Chemical structure of PEO PDMS additives a IM22 m n 15 b Silwet 46Figure 4 6 Phase images of Pebax 1657 blend membranes obtained from SPM 48Figure 4 7 SPM phase images of Pebax 1074 blend membranes 48Figure 4 8 CO2 permeability a and selectivity b of Pebax 1657 blend membranes with 10 50 wtIM22 and Silwet conducted 35 C and 4 bars The dash line indicated the theoretical prediction of CO2permeability contributed by the miscible PEO in IM22 and Silwet co polymer 51Figure 4 9 CO2 solid line and N2 dash line permeability of Pebax 1074 blend membranes with10 40wt IM22 and Silwet 35 C and 200psi 51Figure 4 10 CO2 permeability a and CO2 N2 selectivity b of Pebax 1657 blend membranes with20 IM22 30 and 50 Silwet at different measurement pressure operated at 35 C 52Figure 4 11 CO2 permeability a and CO2 N2 selectivity b of Pebax 1657 blend membranes with 20IM22 and 30 Silwet in pure gas dash line and CO2 N2 mixture solid line as a function of pressuretested at 35 C 53Figure 4 12 SEM image of the PES composite membrane with PEBAX as the coating layer 54Figure 4 13 Schematic representation of the dip coating facility and the sandwich like coating layerstructure of a composite membrane 55Figure 4 14 Relative silicon concentration profile within 10 m depth obtained from EDX analysis forTFC membranes coated with two gutter layers 57Figure 4 15 Chemical structure formular of PTMSP 123 58Figure 4 16 SEM outer cross section images of TFC membranes with PVDF substrate 59Figure 4 17 SEM images of out surface of PVDF hollow fiber substrates coated with PDMS andPTMSP with 2 and 4 layers of coating 59Figure 4 18 CO2 permeance and CO2 N2 permselectivity of TFC membrane measured during theextensive period room temperature 61Figure 4 19 High performance TFC membranes reported in literatures hollow fibre 31 52 125126 flat sheet 4 34 48 50 108 127 this study at room temperature the grey frame indicatesthe target region defined by MTR for high performance TFC membrane 62Figure 5 1 Schematic representation of the membrane permeation set up for the pure and mixed gasas well as for water vapour tests 63Figure 5 2 Comparison of the CO2 N2 gas separation performance with without NO from Matrimidhollow fibre membranes with 4 Silwet L 7607 The numbers show on the top of each column arethe actual permeance and selectivity with without NO A CO2 permeance and B CO2 N2 selectivityFigure 5 3 Simulation results of the CO2 concentration in permeate as function of CO2 N2 selectivityCO2 permeance is assumed at 20 GPU 67Figure 5 4 Schematic representation of competitive sorption caused by water vapour 68Figure 5 5 The effect of temperature and water vapour activity on separation of humidified CO2 69Figure 5 6 The effect of water vapour in the feed evaluated with gas mixture The legends in bothfigures are the same 70Figure 5 7 Comparison of the CO2 N2 gas separation performance with pure gas and mixed gasCO2 N2 20 80 vol under 35 C dash line pure gas solid line mixed gas 71Figure 5 8 Comparison of the CO2 N2 gas separation performance with NO and water under 35 Cdotted line without NO solid line with NO dashed line with water vapour water activity between 0 1to 0 17 73Figure 6 1 Legend for the P ID and PFD 80Figure 6 2 P ID of the membrane capture pilot plant at he Vales Point Power Station 81Figure 6 3 PFD and stream tables for the membrane capture pilot plant at the Vales Point PowerStation 82Figure 7 1 Overall dimensions of the mobile membrane unit 84Figure 7 2 Sampling side of the mobile membrane unit 85Figure 7 3 The system control side of the mobile membrane unit 85Figure 7 4 Floor plan for the CO2CRC mobile membrane unit at Vales Point 87Figure 7 5 CO2 permeance profiles over 3 days on site operation 91Figure 7 6 CO2 N2 selectivity profiles over 3 days on site operation 91Figure 7 7 Permeation flux profile of Module 2 over 3 days operation 93Figure 7 8 Modification of flue gas in let pipe connection 94Figure 7 9 Modifications of the on site membrane unit for the 2nd generation membrane testingupper gas flow rate monitor and lower 2 L water trap 95Figure 7 10 The membrane performance profiles for the composite hollow fiber or 2nd generationmembrane over 17 days on site operation module 1 upper permeance and permeation flux lowerCO2 N2 selectivity 97Figure 7 11 The 2nd generation membrane performance profiles over 17 days on site operationmodule 1 3 upper CO2 permeance lower CO2 N2 selectivity 98Figure 7 12 Corrosion of the copper membrane module fitting after membrane flooding 1001 IntroductionDescription of the projectThis research aimed to fabricate high performance hollow fiber membranes for CO2 capturefrom flue gas and to compare their laboratory performance using synthesised gas mixtureswith real flue gas streams in power plants and to demonstrate the feasibility of membraneapplication for CO2 capture in black coal fired post combustion flue gas Through this studyissues related to translating lab performance to industrial applications such as the influenceof pre treatment processes the minor components in the flue and effects of long termexposure are highlighted The outcomes of this project contribute to advancing technologydevelopment for successful demonstration of low emission coal technology in Australia byidentifying and developing appropriate membrane materials and optimising future processconfigurationsThe challenges for post combustion capture include maintaining acceptable CO2 N2separation while achieving ultra high permeabilities for low feed pressures and relatively lowCO2 concentration in the feed Recent research and development around the world havereported the potential of new polymeric materials and modules in gas separation membranesystems to achieve significantly higher CO2 permeance by two to three orders of magnitudecompared to conventional gas separation membranes In conjunction with smart design ofmulti cascade systems it is anticipated that the costs of CO2 capture from flue gas usingmembranes could be significantly reduced to provide a significant cost advantage with muchfewer environmental issues compared to MEA solvent adsorption processesThis research study utilised our capacity in fabricating polymeric hollow fibres membraneswith high selectivity and our extensive experience in evaluation and understanding theperformance of membrane systems in gas separation applications We have also embarkedon development of new generation composite hollow fiber membranes with high permeationrate to handle post combustion flue gas CO2 capture in this project Hollow fibres providehigh surface area and flexible module configurations adaptable to a number of flue gasseparation processes and rapid scale upThe project was defined into three phases