Lecture 25 2023: CO2 Capture: Oxy-fuel Processes PDF

CO2 Capture: Oxy-fuel Processes Carbon Capture during Power Generation Fuel N2 CO2 Separation Power Generation Post Combustion CO2 Air H2O Air/O2/Steam Air Pre-combustion Fuel Gasification Reformer + CO2 Separation Power Generation N2 H2O CO2 H2O Fuel Power Generation CO2 H2O Air...

CO2 Capture: Oxy-fuel Processes Carbon Capture during Power Generation Fuel N2 CO2 Separation Power Generation Post Combustion CO2 Air H2O Air/O2/Steam Air Pre-combustion Fuel Gasification Reformer + CO2 Separation Power Generation N2 H2O CO2 H2O Fuel Power Generation CO2 H2O Air O2 Separation N2 Oxy-fuel Oxy-fuel Technology Reasons for consideration: • Eliminates N2 from combustor outlet simplifying the carbon capture step. • CO2 capture cost is potentially competitive with other emergent technologies. • Technology is relatively simple although still under development • Significant reduction in NOx Typically, the optimum O2 concentration from the ASU* is around 97-98% with an optimum flue gas (primarily CO2 and H2O) recirculation rate of about 70% to yield a 25-30% O2 to the boiler. Under these conditions the flame and heat transfer characteristics approximate those for air-fired boilers. *ASU = air separation unit Adiabatic flame temperature calculation To illustrate this last point, consider the simple case of coal combustion (assumed here to be primarily solid carbon). Under adiabatic, isobaric conditions the total enthalpy change of the stoichiometric combustion process C(s) + O2(g)  CO2(g) is, assuming the reaction goes to completion, H T = 0 = nI ( H I (T f ) − H I (To )) + nCO2 ( H CO2 (T f ) − H CO2 (To )) + nCO2 H r (To ) Further simplifying by assuming the gas phase is ideal and that H i (T f ) − H i (To ) = C piig (T f − To )  T f = To − nCO2 H r (To ) ig nI C pIig + nCO2 C pCO 2 Writing this as H r (To ) T f = To − ig (nI / nCO2 )C pIig + C pCO 2 Case 1: Pure O2 feed to the combustion process (nI = 0) With To = 298K, H r (To ) = − 393.509 kJ/mol ig then by trial and error C pCO = 70.29 J/mol.K and T f = 5,598 K 2 Case 2: Air feed to the combustion process (I = N2, nI/nCO2 = 3.76) ig ig C pCO = 54 . 6 J/mol.K and C pN2 = 33.4 J/mol.K 2 T f = 2,184 K Case 3: Recycle CO2 and O2 feed to the combustion process (I = CO2,) In this case to obtain Tf = 2,184 K with recycled CO2 at 298K nCO2 (recycle)/nCO2(reaction) = 2.82 (i.e. 26.2% O2 in feed) Challenges 1. Investigation of heat transfer and flue gas clean-up for retrofit systems and the development of novel designs for new installations. 2. Heat recovery is significant to improving the efficiency of oxy-combustion systems. 3. Oxy-fuel power cycles would benefit from fundamental data on oxy-fuel gas-phase flame properties (e.g. flame speeds and radiant heat transfer). 4. Development of an understanding of the character and distribution of ash and slag in pulverised coal oxycombustion systems. 5. Design and optimisation of the air separation unit. Sourcing O2: Air Liquifaction - Preamble Air Separation using Liquefaction At this time, separation of air by distillation is the preferred option for large-scale Oxygen requirements. The steps involved include: • Air compression; • Heat exchange to cool air for liquefaction; • Refrigeration to allow liquefaction; • Distillation to separate the components in air (mainly O2, N2 and trace Argon (~1%)) – operating pressures of between 1-20 bar. • Plants produce >100 tonnes O2/day. To operate Moneypoint on the basis of oxy-fuel technology will require ~ 19,000 tonnes O2/day*. * The current maximum tonnage plant has a capacity of 4,500 (W.F. Castle, Int. J.Ref. 25 (2002) 158-172). Plate or Packed Column Distillation Condenser qC T-x-y plot for N2/O2 (PRO II – Peng Robinson) Vapour Liquid N2 product Reflux (>99.5%) Feed Purified Air -181.50 D B D B D D B D D -184.50 BB D B B D B D B D B D B D B D B D B D BB DD B D -187.50 BB DD BB DD B D BB DD BB DD BB DD BB DD BB DD BB DD BB DD -190.50 BB DD BB DD BB DD BB DD BB D DD BB DD BB DDD BB BBB DDD BB DD BBB D DD BBB D -193.50 DD BBB DDD BBB D DD BBB D DD BBB DD D BBB DD D BBB DD BBB D D D BBB D DD D BBD DB B D Temperature, C Vapour T vs y T vs x Vapour Liquid Reboiler qB Liquid -196.50 0 O2 product (>99.5%) 0.2 0.4 0.6 x* 0.8 Mole fraction (x or y) of N2 (p = 1 bar) y* 1.0 Example y1, V 1 M1 (Air) x1, L 2 3 -181.50 y2, V y3, V M3 y4, V x3, L x1L+y3V = x2L+y2V zM2 = (x1L+y3V)/(L+V) The case shown in the figure is for L=V D B D B D D B D D -184.50 BB D B B D B D B D B D B D B D B D B D BB DD B D -187.50 BB DD BB DD B D BB DD BB DD BB DD BB DD BB DD BB DD B DD -190.50 BB DD BB DD BB DD BB DD BB DD BB DD BB DD BBB D DD BBB DDD BB DDD BBB DD BBB -193.50 DD BBB DDD BBB D D D BBB DDD BBB DD D BBB D D DD BBB DD BBB D D BBB D DD D BBD DB M1 B D M2 M3 Vapour M2 x2, L T-x-y plot for N2/O2 (PRO II – Peng Robinson) Temperature, C x0, L T vs y T vs x Liquid -196.50 0 x3 0.2 z z z x2,y3 0.6 x1,y2 0.8 Mole fraction (x or y) of N2 (p, = 1 bar) y1 1.0 Air Separation using Membranes N2-rich exhaust gas O2- rich gas O2 N2 -rich interior N2 O2-rich gas Air Ls N2, O2 O2 L Ls O2 N2, O2 Polymer or Oxygen Transport Membrane Composite Membrane Microporous support L substrate O2 Hollow tubular porous membrane element Nanoporous thin film or Oxygen Transport Membrane ‘Solution-diffusion’ mechanism Membrane Module Design For a permeance independent of concentration the flux of component i through the membrane is N i = Pi ( pTH yiH − pTL yiL ) where Pi is the permeance (DiKi/LsRT with Di = diffusivity and Ki = adsorption or solubility coefficient), pT and y are the total pressure and gas phase mole fraction and the subscripts H and L refer to the high pressure and low-pressure sides of the membrane respectively. Permeate gas FTL,OUT Feed, FTH,IN FTL,OUT Retentate FTH,OUT Ls FTH,IN Mixed NT FTH,OUT FTH FTH + dFTH The module to be considered is modelled as a cross-flow system with plug flow within the tubes (high pressure side)*. *The more complex situation of plug flow on both sides is analysed by Davis (2005). N2-rich exhaust gas O2- rich gas Air FTH,IN, y1H,IN y For separation of a binary 1-2 gas mixture, the mass transfer area, A, of the membrane module is a primary design variable. Differential analysis of the changes in flow and concentration along the flow-path due to permeation through the membrane as indicated in the previous slide provides y 1H 1H  FTH  1  =− ln  dy1H = −  g ( y1H ,  ,  )dy1H   F y − y 1L y1H ,IN 1H y1H ,IN  TH , IN  y1H ,OUT A = A*  y1H ,IN g ( y1H ,  ,  ) = FTH g ( y1H ,  ,  )dy1H FTH , IN ( y1H − y1L ) + (1 /  )( y2 H − y2 L )   P p  = 1 ;  = TL P2 pTH  FTH , IN    A * =  P1 pTH   1 2  1   1  1    y1H +   − y1H −   + 4 y1H  + − y1H −   2   1 −  1− 1−       Special cases: (1) If the pressure is low on the permeate side ( → 0) then it may be shown that the ratio of the total molar flowrates of the retentate relative to the feed is FTH ,OUT FTH , IN  y1H ,OUT   =    y1H , IN   1      −1   y2 H ,OUT     y   2 H , IN        1−  = 1− FTL ,OUT FTH , IN The mass transfer area of the membrane may also be shown to be given by A = A*  ( − 1) y1H , IN y2 H , IN  y1H   y y1 H ,OUT  1H , IN y1 H ,IN      2 −      −1   1 − y1H  1− y 1H , IN       2 −1     1−  dy1H Note that once the selectivity is specified and the degree of recovery is fixed, the membrane area scales through A* as the inverse of the high-pressure limit and the permeance of the desired component P1. Robeson plot (Robeson (2008)) for the relationship between the O2/N2 selectivity and O2 permeability for polymeric membranes. (Note: 1 Barrer = 2.987x1015 x L x Pi (mol/m2.Pa.s) (2) If the membrane is semi-permeable to component 1 ( → ) then it may be shown that the ratio of the total molar flowrates of the retentate relative to the feed is simply FTH ,OUT FTH , IN and  1 − y1H , IN  FTL,OUT   =  = 1−  FTH , IN  1 − y1H ,OUT   1  ( y1H ,OUT − 1)( y1H , IN −  )   A = A * (1 − y1H , IN )  ln  2   (  − 1)  ( y1H ,OUT −  )( y1H , IN − 1)   1  1 1  + − (  − 1)  ( y1H , IN − 1) ( y1H ,OUT − 1)  The above results may be applied to the design of oxygen ion transport membranes (oxygen-deficient perovskite based ceramic membranes) which are a developing field of research for oxy-fuel applications. Perovskites for Oxygen Transport Membrane (OTM) Materials: These materials usually consist of rare-earth, alkaline-earth and transition metal elements and have the general formula ABX3 'A' and 'B' are two cations of very different sizes, and X is an anion (oxygen in the present case) that bonds to both. Typical oxygen-deficient oxides for use in OTMs include La0.7Sr0.3Ga0.6Fe0.4O3-; Ba0.5Sr0.5Co0.8Fe0.2O3-; Ba1.0Co0.7Fe0.2Nb0.1 O3- The membranes need to be operated at high temperatures (greater than 1200K) to ensure relatively high effective permeances and the latter can be of order 10-8 - 10-7 mol/m2.Pa.s OTMs operate in three stages. 1. O2 is adsorbed from the gas mixtures and split on the oxygen vacancies on the membrane surface (O2+4e- → 2O2-); 2. Oxygen bulk diffusion across the membrane; 3. Surface reaction between lattice oxygen and electron-hole at the surface on the low pressure side (2O2-+4(e-+h+)→ O2+4e-). With sufficiently thin membranes step 2 offers negligible mass transfer resistance (Liu et al (2006)) leading to O2 permeance ~ 5 x 10-8 mol/m2.Pa.s under normal operating conditions. Disadvantages: High operating temperatures (~1200K) requiring careful consideration of housing construction and maintenance Much work is still needed to ensure that only the surface reaction controls i.e. the diffusion limitations of the perovskite and the supporting membrane must be considered. Oxygen Transport Membrane Module Design Permeate gas FTL,OUT (specified by application) 4 Retentate FTH,OUT 3 Feed, 21% O2 FTH,IN mol/s (to be determined) Note that A* = FTH,IN / (P1pTH) and given FTL,OUT then the feed flow requirement is obtained from A/A* pTL/ pTH = 0.1 2 0.06 1 0.03 0.01 FTH , IN (1 − y1H ,OUT ) FTL,OUT = ( y1H , IN − y1H ,OUT ) 0 0.08 0.12 0.16 yO2H,OUT 0.2 0.24 Membrane area required for oxygen separation using OTMs at four pressure ratios Case Study 1: Oxy-fuel operation in an automobile: Typically, an automobile running at 60 km/hr requires 4litre/hr or 0.007mol/s of fuel (modelled as isooctane C8H18) requires a minimum 0.0875 mol/s O2 for combustion C8H18 + 12.5O2 → 8CO2 + 9H2O With a feed air pressure of 10 bar and 21% Oxygen and an OTM permeance of 5 x 10-8 mol/m2.Pa.s then if the outlet air composition on the high pressure side is 10.5% Oxygen, the feed air requirement is 0.746 mol/s and A* = 14.92 m2. From the graph we have the following results.  0.1 0.06 0.03 0.01 A (m2) 49 19 14.5 12 Case Study 2: Oxy-fuel operation in Moneypoint (coal-fired): The CO2 effluent is 6.77x103 mol/s = O2 flowrate*. Hence the air feed is 57.7x103 mol/s  0.1 0.06 0.03 0.01 A (m2) 3.9x106 1.47x106 1.12x106 0.93x106 *Current membrane systems are limited to < 400 tonnes O2/day (145 mol/s) (Castle (2002)). Air Feed N2 - rich Win(Net) Turbine Membrane Module Compressor Permeate (O2 ) Energy Costs: To combustion with recycled CO2 The membrane module operates at ~ 1000oC and the incoming air is to be compressed to 10 bar ( = 0.85). Thermal energy is recovered from the hot Nitrogen exhaust gas by interchange with the incoming air and the compression energy is partially recovered during decompression of the Nitrogen rich exhaust stream from 10 bar to 1 bar. Per mole of feed air the energy penalty for the cases outlined earlier is approximately 30%. Minimum Power Requirements for Oxy-fuel Combustion CO2 Capture (CH4 as Fuel) CH4 Power Generation CO2 2H2O 2O2 Air ASU N2 Based on total CO2 emissions of 2,470 kmol/s with fully reversible separation of air ( yO2 = 0.209) producing pure O2 and compression of the pure CO2 product to 100 bar at 298K 1  p    y 2 X ln y X  RT 2 FCO2 (GW ) Minimum Power =  ln   −  ln yO2 +   yO2  2  p1    = (27.6 + 30.7) = 58.3 GW ( ) EU power demand (total) = 2320 GW EU electric power demand ~ 450 GW Reasons for consideration: • Eliminates N2 from combustor outlet simplifying the carbon capture step. • CO2 capture cost is potentially competitive with other emergent technologies. • Technology is relatively simple although still under development • Significant reduction in NOx