Salinity gradient energy: Assessment of pressure retarded osmosis and osmotic heat engines for energy generation from low-grade heat sources

Transcription

1 Engineering Conferences International ECI Digital Archives Advanced Membrane Technology VII Proceedings Salinity gradient energy: Assessment of pressure retarded osmosis and osmotic heat engines for energy generation from low-grade heat sources Tzahi Y. Cath Colorado School of Mines, Johan Vanneste Colorado School of Mines Michael B. Heeley Colorado School of Mines Kerri L. Hickenbottom Humboldt State University Follow this and additional works at: Recommended Citation Tzahi Y. Cath, Johan Vanneste, Michael B. Heeley, and Kerri L. Hickenbottom, "Salinity gradient energy: Assessment of pressure retarded osmosis and osmotic heat engines for energy generation from low-grade heat sources" in "Advanced Membrane Technology VII", Isabel C. Escobar, Professor, University of Kentucky, USA Jamie Hestekin, Associate Professor, University of Arkansas, USA Eds, ECI Symposium Series, (216). This Abstract and Presentation is brought to you for free and open access by the Proceedings at ECI Digital Archives. It has been accepted for inclusion in Advanced Membrane Technology VII by an authorized administrator of ECI Digital Archives. For more information, please contact

2 Salinity gradient energy: Assessment of pressure retarded osmosis and osmotic heat engines for energy generation from low-grade heat sources Johan Vanneste, Kerri L. Hickenbottom, Tzahi Y. Cath Colorado School of Mines Dept. of Civil and Environmental Engineering ECI, Advanced Membrane Technology VII September 14 th, 216 Cork, Ireland

3 Structure of the Presentation Low-grade heat Membrane Distillation Pressure Retarded Osmosis Output: Energy 2

4 Structure of the Presentation Part I: Pressure retarded osmosis (PRO): membranes and spacers evaluation Part II: Osmotic heat engine (OHE): working fluid selection Low-grade heat Membrane Distillation Pressure Retarded Osmosis Part III: Technoeconomic assessment of the OHE Output: Energy 3

5 Pressure Retarded Osmosis (PRO): Utilizing Energy from the Ocean Draw Solution High Concentration High Pressure Feed Solution Low Concentration Low Pressure PX Turbine Semi-permeable, hydrophilic membrane Applied Pressure < Osmotic Pressure Water Flux (J w ) = A(Δπ-ΔP) (simplistic!) Power Density (W) = J w ΔP (power output per membrane area) 4

6 Open-Loop PRO Applications Seawater River water RO concentrate freshwater Dead Sea and Great Salt Lake Limitations of open loop PRO Extensive pretreatment needed Potential irreversible membrane fouling Solution chemistry and temperature are fixed Source: Statkraft 5

7 Closed-Loop PRO: Osmotic heat engine (OHE) Controlled solution temperature and chemistry High osmotic pressures yields high power densities Lower reverse salt flux Can utilize low-grade heat Potential for energy storage Closed loop system = No backwashing or chemical cleaning! Thermal separation LC J w HC W h,pro W t HC = high concentration (DS) LC = low concentration (DI water) 6

8 Membrane Distillation (MD): Utilizing LGH for Draw Solution and Feed Stream Regeneration High Concentration High Temperature Low Concentration Low Temperature J w Microporous, hydrophobic membrane Water Flux (J w ) = A w (ΔP v *) (simplistic!) Can utilize low-grade heat to simultaneously separate and concentrate mixed streams 7

9 PRO MEMBRANE AND SPACER ASSESSMENT 8

10 PRO: An Energy Generating Membrane Process PRO membranes are not yet commercially available and FO membranes are used PRO membranes must: J s Porous Support ICP C Active Layer C D,m J w C D,b Exhibit high power density & low reverse solute flux C F,b C F,m ECP D LC HC Withstand high operating pressures PRO membrane spacers must provide good mixing and adequate support HC = high concentration LC = low-concentration C D,b = bulk draw concentration C D,m = membrane interface draw conc. C F,b = bulk draw concentration C F,m = membrane interface draw conc. ECP dil = dilutive external CP ICP C = concentrative external CP 9

11 Operating Conditions: PRO Source water Draw solution (DS): 1, 2, & 3 M NaCl Feed: deionized water Bench scale testing SCADA system controls temperatures, and collect data to calculate water flux, batch recovery, and salt rejection Flat sheet FO membranes Hydration Technologies Innovations (HTI) thin-film composite (TFC) HTI cellulose triacetate (CTA) Oasys Water TFC X company TFC 1

12 PRO Bench-Scale System Membrane module PROzilla Membrane spacers Tricot (35-ch) Tricot (2-ch) Extruded mesh 11

13 Specific reverse solute flux, g/l PRO Membrane Evaluation: DS hydraulic pressure, psi DS hydraulic pressure, psi Power density, W/m DS hydraulic pressure, MPa DS hydraulic pressure, MPa X HTI CTA Oasys HTI TFC X HTI CTA Oasys HTI TFC 1 M NaCl draw solution (DS); 2x 2 channel tricot on feed High power densities Low specific reverse solute flux Good mechanical stability 12

14 Spacer Orientation Feed spacer orientations 25 DS hydraulic pressure, psi Parallel to flow Power density, W/m M 2 M 2 48 % increase! 3 M to flow DS hydraulic pressure, MPa HTI TFC 3 M HTI TFC 2 M HTI TFC 1 M Corrugated heat exchanger plate Spacerless patterned membranes! 13

15 Water flux, L m-2 hr-1 Membrane Deformation Surface Crosssection Virgin membrane Membrane after 45 psi Membrane after 6 psi Specific reverse solute flux, g L Js/Jw Jw 7 psi Elapsed time, hr Importance of long term experiments!!! 14

16 OHE WORKING FLUID SELECTION 15

17 Reverse Solute Flux in PRO and Implications for OHE Process Performance Non-idealities in PRO gives rise to reverse solute flux (RSF, J s ): J s = B π D,b exp J w k 1 + B J w exp J w S D π F,b exp J w S D exp ( J w k ) J w = A π D,b exp J w k 1 + B J w exp J w S D π F,b exp J w S D exp ( J w k ) P HC J w LC LC J w HC To sustain osmotic driving force, must bleed a portion of the PRO feed stream to the MD feed MD J s PRO Impacts of RSF on net power outputs and generation costs are unknown! W t Legend: π = osmotic driving force k = mass transfer coefficient (DS side) S = membrane structural parameter B = solute permeability coefficient D = draw solution diffusivity 16

18 LiCl LiBr CaCl2 MgCl2 Na(C2H5COO) HCOONa NaCl KBr (NH4)2SO4 KCl C2H3NaO2 Na3C6H5O7 Na2SO4 Mg(CH3COO)2 MgSO4 NH4HCO3 KHCO3 NaHCO3 Osmotic pressure, MPa PVP, kpa Draw Solution Screening and Selection 6 Osmotic pressure PVP (b) 24 Draw solution Cost Concentration 5 2 $/kg M 4 16 CaCl HCOONa KBr LiBr LiCl MgCl Na(C 2 H 5 COO) NaCl 6 3. Eight salts met screening criteria (π > π NaCl & non-hazardous) Tested at concentrations equivalent to 17.4 MPa osmotic pressure (osmotic pressure of 3 M NaCl) 17

19 Operating Conditions: PRO Source water Draw solution: varied, 3 ºC Feed: deionized water, 3 ºC C P Membrane cell P C Membrane: HTI TFC Draw solution hydraulic pressure kept constant at 2 MPa (3 psi) Dosing ~ Draw Draw loop P T P T Feed loop ~ Feed Control system Legend ~ C P T Ultrasonic sensor Conductivity probe Pressure transducer Temperature probe Flow meter Back pressure valve Gear Pump Peristaltic pump Heat exchanger Plunger pump 18

20 Operating Conditions: MD Source water Feed solution: varied Distillate: deionized water C T Membrane cell T C Stream temperatures Feed solution: 55 ºC Dosing Feed loop Distillate loop Distillate: 25 ºC Flat sheet, hydrophobic, microporous (.2 μm) membranes from 3M Feed Legend P T Control system P T Distillate. ~ Ultrasonic sensor Flow meter C Conductivity probe Gear Pump P Pressure transducer Peristaltic pump T Temperature probe. Scale Heat exchanger Stir plate 19

21 Diffusivity, (1-9 ) m 2 /s PRO Performance Power density, W/m Power density Diffusivity Difference in power densities is because of the difference in diffusivity and solute permeability coefficient (B) 2

22 Reverse solute flux, g/m 2 -h B, L/m 2 -h Diffusivity, (1-9 ) m 2 /s PRO Performance Solute flux Diffusivity B Higher diffusivities lead to higher reverse salt fluxes Salts with high reverse salt fluxes can be detrimental to system costs 21

23 Water flux, L/m 2 -h ΔPVP, kpa MD Performance Flux ΔPVP Distillate conductivity decreased over time, indicating 1% rejection of salts Water flux decreases with decreased partial vapor pressure difference 22

24 Preliminary Economic Evaluation of Draw Solutions Experimental PRO and MD results were used in the model Design constraint: PRO feed concentration of 4 g/l (DS specific) C o,h,md Q o,h,md module costs for PRO MD and C o,l,md Q o,l,md MD C o,h,md Q o,h,md Q i,l,pro C o,h,md Q o,h,md C o,l,md C i,h,md Q o,l,md Q i,h,md C o,l,md Q o,l,md MD PRO Legend J w PRO Q = flow J w C = concentration HC = high concentration HC LC LC = low HC concentration C i,b J w = water flux J s = salt flux C i,l,md Q i,l,md MD-LC Tank W l,md C i,l,pro C i,h,t Q i,l,pro Q i,h,t Legend PRO Q = flow C o,b J C Buffer = concentration J w w C Q o,b o,h,t W TANK h,md HC = high concentration Q o,h,t LC = low concentration HC LC J LC C w = water HC flux i,b J s = salt flux Legend PRO C o,l,md Q = Cflow i,l,pro W l,md Q C = concentration o,l,md C i,h,md C i,l,md Q i,l,pro C HC = high concentration i,h,t Q i,h,md Q i,l,md LC = low concentration Q i,h,t J J J C w = water flux w w i,b J s = salt flux Q i,b W Buffer HC LC h,pro C o,h,t W LC TANK HC h,md Subscripts C i,l,pro C Q i,h,pro i = inlet o,h,t o = outlet MD-LC Tank C o,h,md Specific membrane and Q o,h,md MD were referenced from RO and MF literature, respectively Q Q i,b Q r,l,pro MD PRO-LC Tank MD-LC Tank W h,pro J s Q i,b Q r,l,pro C i,h,pro Q i,h,pro C o,b Q o,b PRO-LC Tank C o,l,pro Q o,l,pro J s C i,b Q i,b Q r,l,pro PRO-LC Tank Subscripts i = inlet o = outlet C o,h,pro r = return Q o,h,pro l = low concentration h = high concentration p = permeate W T b = bleed t = tank T = turbine W l,pro W h,pro C i,h,pro Q i,h,pro C i,l,pro Q i,l,pro C o,l,pro Q o,l,pro W l,pro W h,pro J s C i,h,pro Q i,h,pro Subscripts C i = inlet o,h,pro o = outlet Q o,h,pro r = return l = low concentration W T h = high concentration p = permeate b = bleed t = tank T = turbine Symbols Heat exchanger Gear pump Plunger pump Turbine-generator Pressure exchanger Legen Q = flo C = co HC = LC = l J w = w J s = sa Subsc i = inle o = ou r = ret l = low h = hig p = pe b = ble t = tan T = tu 23 Symb

25 Electricity generation cost, $ / kwh Impact of Draw Solutions on System Costing CaCl 2 and MgCl 2 were the salts that performed best in MD and had the lowest specific reverse solute fluxes Electricity generation costs are higher than expected because of low PRO operating pressures (low power densities) 24

26 Capital cost, $ million Net power, kw System Costing and Net Energy (1 MW gross) Salt costs PRO membrane costs MD membrane costs (bleed only) MD membrane costs (no bleed) Net power MD membranes are the highest costs Salts with high RSF result in more bleeding, subsequently decreasing net energy and increasing MD membrane area 25

27 Water flux, L/m 2 -hr RSF, mmol/m 2 -hr Mixed draw solutions: best of both worlds? Water flux Solute flux Percent MgCl 2 -OP With addition of MgCl 2, high water flux of NaCl can be maintained while RSF drops significantly 26

28 OHE TECHNO-ECONOMIC ASSESSMENT 27

29 OHE System Model and Cost Assumptions Base case scenario 2.5 MW (net power) system Draw solution: 3 M NaCl PRO operating pressure: 3.4 MPa (~5 psi) MD temperatures: 7 C feed, 3 C distillate Assumptions Low grade heat and cooling is free Membrane costs referenced from commercially available membranes PRO costs referenced from RO MD costs referenced from MF Equipment efficiencies Pump efficiency 7 % Turbine efficiency 9 % Pressure exchanger efficiency 95 % Generator efficiency 95 % Heat exchanger efficiency 6 % Data and other assumptions Plant life 2 yr Plant availability 9 % PRO membrane replacement 1 %/yr MD membrane replacement 1 %/yr Interest (discount) rate, I 8 % Inflation rate, n 3 % Amortization factor.1 Labor cost.3 $/m 3 Specific membrane costs * PRO membrane element cost 11 $/m 2 MD membrane element cost 24 $/m 2 PRO membrane housing cost 17 $/m 2 MD membrane housing cost 14 $/m 2 *Assumed 35% mark-down from quoted distributor cost 28

30 OHE Power Generation Power outputs Gross Power: 4.9 MW Net Power: 2.5 MW About 2 % of capital costs due to bleeding Process efficiency Theoretical efficiency 1 4% System efficiency.1% System costs: $.48 per kwh Benchmark <$.2 per kwh Generation Regenerati Control Hydraulic Civil work system on system - PRO 36% - MD 21% (total) 13% 4% 26% Hydraulic system 4 % Cost PRO MD of membrane labor and replaceme spare 82% nt 14% PRO 4% Control system 21 % Capital Cost $57 million Civil work 36 % MD 26 % PRO 13 % MD 14 % Labor 82% O&M Cost $72 million 1 Lin et al., ES&T 214 Hickenbottom et. al, in Prep. 29

31 Generation costs, $/kwh Generation costs, $/kwh Cost Sensitivity: PRO Power Density and solute permeability(b) (29) (58) (87) (1,16) Power density, W/m B, L/m 2 -hr Power density, W/m 2 PRO hydraulic pressure, MPa (psi) Generation cost - w/bleeding PRO power density 3

32 Assumptions and Model Outputs: Ideal Case OHE 25 MW (net power) system (economy of scale) Draw solution: 3 M NaCl PRO operating pressure: 7.6 MPa (1,1 psi), corresponding to a PRO power density of 76 W/m 2 MD temperatures: 85 C feed, 15 C distillate OHE Sensitivities Base Ideal Unit System size (net power) MW Electricity generation cost.48.1 $/kwh PRO operating pressures MPa PRO power density W/m 2 PRO recoveries 15 4 % MD feed (LGH) temperature 7 85 (95) C MD distillate (cooling) temperature 3 15 (5) C MD recoveries 6 3 % MD water flux L/m 2 -h Membrane replacement 1 5 %/yr 31

33 Generation costs, $ per kwh Costs of Competing Energy Generation Technologies $.48 Stationary Engines $.32 $.2 OHE Organic Rankine Cycle (ORC) $.16 $.13 $.13 $.12 $.1 $.1 $.1 $.1 Conventional coal-fired power ($.3 per kwh) $.8 $.9 $.7 *Figure adapted from US EIA s Annual Energy Outlook 212; and ElectraTherm

34 Conclusions HTI TFC membranes are the best commercially available PRO membranes, withstanding operating pressures up to 5 psi and attaining power densities up to 22 W/m 2 Low-grade heat Membrane Distillation Pressure Retarded Osmosis Use of CaCl 2 and MgCl 2 as working fluids, can decrease OHE electricity generation costs by > 46%. Mixed draw solutions with low RSF have the potential to further decrease costs. Changing spacer orientation from parallel to 45 increased PRO power densities with 48 % Output: Energy Future improvements to PRO membranes and power densities (> 76 W/m 2 ) could make the OHE a competitive renewable energy and energy storage technology. 33

35 $/m3 The future of osmotic power and the OHE? 2.5 RO Desalination Costs Year Greenlee et al., Water Ressarce 43 (29)

36 Industry partners Hydration Technologies 3M Academic collaborators Yale University Funding agencies ARPA-E EPA-STAR Research Group Ryan Holloway Mike Veres Tani Cath Estefani Bustos-Dena William Porter Katie Shumacher Curtis Weller Acknowledgements 35

37 Thank you! 36