Next scale chemical looping combustion: Process integration and part load investigations for a 10MW demonstration unit

Available online at www.sciencedirect.com Energy Procedia 37 (2013 ) 635 644 GHGT-11 Next scale chemical looping combustion: Process integration and part load investigations for a 10MW demonstration unit Klemens Marx a *, Otmar Bertsch b, Tobias Pröll a, Hermann Hofbauer a a Vienna University of Technology, Getreidemarkt 9/166, 1060 Wien, Austria b BERTSCHenergy, Josef Bertsch Gesellschaft m.b.h. & Co. KG, Herrengasse 23, 6700 Bludenz, Austria Abstract Chemical looping combustion (CLC) is a second generation carbon capture process and is discussed as a potential breakthrough technology with respect to CO 2 avoidance costs. The potential of CLC has already been successfully demonstrated at scales up to 140 kw power input, using gaseous fuels. The next stage of process evolution is the development of a CLC demonstration plant at industrial scale, to gain sufficient confidence in the technology for further up-scaling. In this work, the integration of a next scale CLC demonstration plant at 10MW fuel power input into a steam generating system of a commercial natural gas combined cycle (NGCC) plant with supplementary firing is investigated. The attached CLC demonstration plant is designed to substitute the energy input of the supplementary firing which in turn can easily compensate fluctuation from demonstration plant operation with very little response time or even increase the overall power output in parallel operation with the CLC plant. Such a system exhibits the advantage that power output from the CLC unit would not contribute to the CO 2 footprint of the site, thus improving the CO 2 output balance. Demonstration plant operation including part load and control behavior is investigated by detailed mass and energy balance investigations. This work can be used as a basis for detailed engineering of a next scale 10MW chemical looping combustion demonstration unit. 2013 The Authors. Published by by Elsevier Ltd. Ltd. Selection and/or peer-review under responsibility of of GHGT GHGT "Keywords: chemical looping; demonstration; scale-up" 1. Introduction Chemical looping combustion (CLC) is a second generation carbon capture process and is discussed as a potential breakthrough technology with respect to CO 2 avoidance costs. CLC is a two step combustion * Corresponding author. Tel.: +43-(0)1-58801-166364; fax: +43-(0)1-58801-9166364. E-mail address: Klemens.Marx@tuwien.ac.at. 1876-6102 2013 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of GHGT doi:10.1016/j.egypro.2013.05.151

636 Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 process where mixing of combustion air and fuel is inherently avoided. This process is generally established by using metal oxides, conducting a continuous reduction-oxidation-cycle to transfer the oxygen needed for combustion between two separate reactors; the air reactor (AR) and the fuel reactor (FR). Hence, combustion flue gases are free from air nitrogen and after water condensation, a CO 2 rich stream is obtained, ideally ready for sequestration (Fig. 1). Therefore, CLC offers a great potential compared to all other carbon capture technologies, as it avoids energy and cost intensive gas-gas separation steps and thus considerable reduces CO 2 avoidance costs. N 2, O 2 CO 2, H 2 O MeO AR FR MeO -1 air fuel Fig. 1 Scheme of a chemical looping combustion process. AR air reactor, FR fuel reactor, Me - metal. The potential of chemical looping combustion for gaseous fuels has been successfully demonstrated in the past in continuously operated dual fluidized bed test facilities, at scales of up to 140 kw fuel input power [1]. By the middle of 2010, more than 4000h of experience in continuous CLC operation had already been collected within the scientific community, including at least eleven chemical looping units [2]. A variety of different oxygen carriers have been tested [3] with nickel based ones being the most intensively studied with more than 1500h of continuous CLC operation [4-5]. Based on previous work of the authors, focusing on reactor design [6], in this work, the integration of a next scale CLC demonstration plant at 10MW fuel power input into the steam generating system of a commercial natural gas combined cycle (NGCC) plant is investigated. Operating experience gained at the 140kW chemical looping pilot plant at Vienna University of Technology is used to determine expectable performance data. Demonstration plant operation including part load and control behavior is investigated by detailed mass and energy balance investigations. This work can be used as a basis for detailed engineering of a next scale 10MW chemical looping combustion demonstration unit. 2. Selection of site and site conditions A key issue for successful demonstration of a new technology is the selection of an adequate site for effective and long term demonstration. To keep the investment costs low integration of the demonstration unit into an existing site is beneficial over green field installation. Utilizing present infrastructures is essential for successful operation within the scope of the selected site. Therefore, the site should allow for integration of the unit at the selected size while keeping the site infrastructure untouched. For this reason a site is selected where a natural gas combined cycle with supplementary firing before the heat recovery

Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 637 section installed is located. The attached CLC demonstration plant is designed to substitute the energy input of the supplementary firing which in turn can easily compensate fluctuations from demonstration plant operation with very little response time. Moreover the overall power output can be increased when operating the CLC plant and supplementary firing in parallel. Such a system exhibits the advantage that power output from the CLC unit would not contribute to the CO 2 footprint of the site, thus improving the CO 2 output balance. supplementary firing steam from CLC-unit generator fuel GT HRSG steam cycle losses generator Fig. 2 Schematic overall process energy flow diagram. GT gas turbine, HRSG heat recovery steam generator. The selected site includes a 40-50 to/h steam production based on natural gas combined cycle. The plant capacity is increased by use of supplementary firing. The plant is a cogeneration plant for heat and power where most of the heat is used for district heating. The CLC demonstration unit shares the steam cycle, including i.e. the steam turbine, condenser and boiler water deaerator, with the existing plant. This allows for significant reduction of the investment costs. A summary of relevant site parameters concerning the steam cycle is given in Table 1. Table 1. Summary of site conditions for the steam cycle Item Symbol Value Unit CLC demonstration plant fuel input power (based on LHV) 10 MW Natural gas combined cycle steam production rate 40-50 to/h Live steam pressure p st 65 bar(a) Live steam temperature st 450 C Boiler feed water temperature BFW 105 C Boiler feed water pressure p BFW saturated water bar(a) 3. Heat and process integration In chemical looping systems the temperature in the reactors is limited by acceptable oxygen carrier operation temperature. Therefore, excess heat needs to be withdrawn from the system which can be done at four different locations: from the exhaust gas streams, from the air reactor, from the fuel reactor, and heat extraction from the circulating solids. The selected system includes an air reactor equipped with water cooled walls and a bed material cooler. While the heat extraction rate from the water cooled wall is governed by riser hydrodynamics (fast fluidized bed) the cooling duty of the bed material (bubbling bed)

638 Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 cooler is controllable by using a solids control flow valve. A schematic of the setup is shown in Fig. 3. More specific details of the reactor system design can be found elsewhere [6]. exhaust exhaust AR FR WW SCV ULS BMC ILS air fuel LLS Fig. 3 Reactor system with heat exchanger arrangement based on the dual circulating fluidized bed concept. Locations where fluidization is applied are indicated by arrows. AR - air reactor, FR fuel reactor, LLS lower loop seal, ULS upper loop seal, ILS internal loop seal, BMC bed material cooler, WW water cooled walls, SCV solids flow control valve. The process flow diagram, shown in Fig. 4, represents the entire chemical looping combustion plant without treatment of fuel reactor exhaust gases. The input streams fuel, air, boiler feed (BFW) water and fluidization steam as well as the output streams superheated steam, AR exhaust gas and FR exhaust gas define the system boundaries of the plant. Air is provided at ambient conditions and is preheated by air reactor exhaust gas (air_ph) for energy recovery after compression (air_blower) to AR inlet pressure. Natural gas is provided at elevated pressure and is used as a fuel. The boiler feed water is supplied from the boiler feed water deaerator and is compressed to boiler pressure by the boiler feed water pump (BFWP). Fluidization steam is taken from an external source, i.e. from a low pressure section of a steam turbine. The process flow diagram covers the CLC boiler and the heat recovery steam generator. A common natural circulation steam cycle is integrated where steam is produced utilizing thermal forces to overcome the pressure drop within the evaporative heat exchanger surfaces. At the operating parameters chosen the amount of heat produced in the reactors is greater than the sensible heat in the exhaust gas streams. Therefore a bed material cooler (BMC) and water cooled walls inside the air reactor (AR_clg) are arranged. Only a part of the elutriated solids from the air reactor are taken and directed to the bed material cooler controlled by a hot sand valve. The setup of the heat recovery steam generator includes the AR and FR heat recovery boiler and the bed material cooler. The proposed AR and FR heat recovery boiler consists of steam super heaters (SH_I and SH_II), economizers (ECO_AR and ECO_FR) and an air preheater (air_ph). To control the steam temperature a drum heat exchanger is integrated. The quantity of steam passing the drum heat exchanger depends on the plant fuel load operating conditions. To overcome

Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 639 WAT superheated steam SH_II ECO_AR FR ID fan SH_I AR_clg ECO_FR GAS FR exhaust AR BMC FR BFWP GAS fuel WAT BFW WAT support steam air_ph GAS AR exhaust AR IDfan AMB air Fig. 4 CLC process scheme. the pressure drop in the heat recovery boiler two induct fans are used; the fuel reactor induct fan (FR ID fan) and the air reactor induct fan (AR ID fan). 4. Process performance In CLC systems the fuel combustion efficiency to CO 2 is subjected to uncertainties. Pilot plant operation experience is used to determine expectable CLC performance at demonstration scale. Anyhow, operating conditions such as gas-solids contact efficiency and contact time may differ significantly from pilot to demonstration scale. For this reason three cases are considered. In the case the combustion performance determined in 140kW pilot benchmark testing is used. In the case the theoretically reachable efficiency is determined by assuming thermodynamic equilibrium in the FR. A third, the case, a fuel conversion is considered representing the average mean, in terms of combustion efficiency, of the case and. The case can be expected to be a meaningful basis for the basic design of such a CLC plant. The considered oxygen carrier is a nickel based oxygen carrier which has been tested for far more than thousand hours [4]. A summary of relevant demonstration plant parameters is given in Table 2.

640 Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 Table 2. Summary relevant demonstration unit parameter Item Symbol Value Unit Fuel reactor temperature FR 900 C Fuel combustion efficiency based on gases LHV: case comb 96.1 % Fuel combustion efficiency based on gases LHV: case comb 99.2 % Fuel combustion efficiency based on gases LHV: case comb 97.7 % Global air/fuel ratio at nominal load 1.1 - Fuel reactor design superficial gas velocity U s_fr 5.5 m/s Air reactor design superficial gas velocity U s_ar 7.5 m/s Estimated total fuel reactor bed pressure drop FR 200 mbar Estimated total air reactor bed pressure drop AR 200 mbar Reactor system heat loss (rel. to total energy input) q loss 1.5 % Blowers isentropic efficiency blower 82.5 % Loop seal steam to air reactor m st_ar 178.2 kg/h Loop seal steam to fuel reactor m st_fr 167.9 kg/h Estimated air reactor solids entrainment flux Gs AR 50 Oxygen carrier definition Redox-system - Ni/NiO Active nickel content - 40.6 % Mean particle diameter d p 140 Apparent density p 3425 kg/m 3-2 s -1 A simple energy flow diagram of the case is shown Fig. 5. The values of sensible heat are based on 25 C, 1.0 bar(a) and gaseous species as a reference. The chemically bound energy is calculated based on the lower heating value. The energy of the water/steam streams refers to the enthalpy of the boiler feed water. This allows direct calculation of the efficiencies from the diagram. The input streams air, fuel support steam and electricity and the output streams stack loss, heat loss and superheated steam are presented. Non utilizable energy is leaving the system via the stack in the form of chemical energy, caused by incomplete combustion, and sensible heat as well as reactor system heat loss. Energy is recovered prior to the stack in an air pre-heater (air_ph). A major amount of the energy input is converted to steam. Further improvement of the process can include reduction of the losses due to unconverted fuel and further utilization of the sensible heat of the exhaust gases. This can include using the heat of condensation of fuel reactor exhaust gas stream. The gas temperature and water/steam cycle temperature profiles for the case are shown in Fig. 6; in a Q-T-Diagram. Conservative pinch point temperatures are assumed considering the scale of the plant. The analysis reveals that more than 50% of the totally transferred heat is withdrawn from the reactor system itself using the bed material cooler and the water cooled walls in the air reactor. The remaining energy leaves the system via the reactor exhaust gas streams. The major part of approximately 52.5% of the energy released is needed to generate steam while only about 20% is needed for superheating. This ratio is typical for the chosen steam parameters. Considering the three fuel conversion cases it turns out that with the proposed setup the boiler efficiency ranges from 88.00 to 94.26% which is the amount of steam produced referred to the fuel input. Considering the small size these values are in typical range; improvement will require increased heat

Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 641 Fig. 5 Energy flow diagram for the case. Indicated in light red is the sensible heat and in yellow the chemical bound energy in the gas streams. Shown in dark red is either the energy entering the system as electrical energy or the energy leaving the system to the surrounding as heat loss. Given in blue is the energy in the water/steam. Reference temperature for calculation of thermodynamic properties is 25 C and 1.0 bar(a). Chemical energy bases on the lower heating value of the gas. exchanger sizes which in turn increase the investment costs. Considering the combustion efficiency scenarios it turns out that a significant amount of fuel cannot be directly used in the CLC system; which is in the worst case closely 7%. Reduction by suitable measures is obligatory. This can include oxygen polishing of fuel reactor flue gases or separation and recycling of combustibles during CO 2 compression and liquefaction [7]. A summary of expectable process parameters for the considered cases is given in Table 3. 5. Part load operation In chemical looping combustion systems the energy to be withdrawn from the reactors has to be adapted accordingly to keep the desired reactor temperatures. The energy extraction rate depends on the reactor temperature, the air/fuel ratio, the combustion efficiency of the process and the total energy input. While the reactor temperature and combustion efficiency is an extensive process parameter the air/fuel ratio can be set arbitrary. In general the energy to be extracted from the reactor system decreases with rising air/fuel ratio and reduced combustion efficiency.

642 Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 1000 1000 AR exhaust 800 circ. solids AR riser FR exhaust 800 Temperature in C 600 400 BMC AR clg SH II SH I 600 400 200 AR ECO air ph FR ECO 200 0 0 0% 20% 40% 60% 80% 100% Relative heat transfer rate Fig. 6 Q-T-Diagram of the proposed setup for the conversion case. Table 3. Summary of expectable process parameters Item Symbol Case Case Case Unit Fuel power P fuel 10 MW Combustion efficency comb 96.1 99.2 97.7 % Fuel reactor temperature FR 900 C Air reactor temperature AR 950 955 965 C Air/fuel ratio 1.1 - Steam production temperature st 450 C pressure p st 65.00 bar(a) Steam flow rate m st 11094.9 11490.5 11884.5 kg/h Electrical power demand P el 110.12 111.02 111.92 kw Boiler efficiency b 88.00 91.14 94.26 % Thermal efficiency th 86.90 90.03 93.14 % Loss due to unburnt fuel (based on LHV) l u 6.97 3.87 0.78 %

Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 643 In the proposed reactor system arrangement the heat extraction rate can be influenced through the air/fuel ratio and through solids flow to the bed material cooler. The solids flow can be controlled using the solids flow control valve. The more solids pass the bed material cooler the higher the heat extracted. By doing so the energy balance can be controlled even in part load operation when the heat extraction rate diminishes. To quantify the amount of solids flowing to the bed material cooler the solids split-up ratio X BMC is defined to (1) X BMC describes the amount of solids flowing to the BMC relative to the total amount of solids entrained from the air reactor passing through the downcomer. Given that proper loop seal operation requires keeping a certain solids level in the loop seal X BMC is limited in practice. Bed material cooler duty in kw 3500 3000 2500 2000 1500 1000 AR=925 C AR=930 C AR=935 C AR=950 C AR=945 C AR=940 C X BMC=20% X BMC=25% X BMC=40% X BMC=35% X BMC=30% AR=955 C AR=960 C AR=965 C AR=970 C AR=975 C X BMC=37.5% X BMC=2.5% AR = 955 C U AR = 7.8 m / s = 1.1 X BMC = 29.9% X BMC=5% X BMC=7.5% X BMC=12.5% X BMC=10% X BMC=17.5% X BMC=15% FR comb Gs AR A AR X BMC=27.5% X BMC=22.5% = 900 C = 97.675% = 50 kg / m²s = 1.44m² X BMC=32.5% 3500 3000 2500 2000 1500 1000 500 (k A) AR A BMC = 8.44 kw / K = 8.96m² 500 0 0 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 Fuel input in kw Fig. 7 Part load diagram considering a combustion efficiency of case. A part load operation diagram is given in Fig. 7. It shows how the necessary heat extraction rate by the solids spilt-up-ratio X BMC and the gas velocity in the air reactor riser U AR at a given fuel reactor temperature FR and process combustion efficiency comb a fuel power input of 10MW the air reactor temperature is expected to be AR 30% (i.e. X BMC ) of the solids entrained by the air reactor have to pass the bed material cooler. The solids split-up-ratio X BMC decreases with excess energy withdrawn by the air reactor exhaust gas; which is

644 Klemens Marx et al. / Energy Procedia 37 ( 2013 ) 635 644 increasing the air/fuel ratio. Furthermore, the part load diagram allows evaluation of operable load conditions within the given parameter setup. 6. Summary and conclusion A heat exchanger arrangement of a possible next scale 10MW fuel chemical looping combustion demonstration plant is proposed and expectable process performance parameters are determined using advanced mass and energy balance investigations. The demonstration plant is considered to be implemented into an existing site of a natural gas combined cycle with supplementary firing. The considered oxygen carrier is based on nickel which has been successfully tested in the past for far more than 1000h in continuous operation. The model within the mass and energy balance investigation structure is validated against performance data obtained in the 140kW chemical looping pilot unit at the Vienna University of Technology. Anyhow, the fuel conversion performance is subjected to uncertainties when going from laboratory to demonstration scale. For this reason, in addition the optimal performance of such a plant is determined by formulating thermodynamic equilibrium in the fuel reactor exhaust. It turns out that the boiler efficiency can reach values of 88-94%. This is in a typical range considering the size of the unit and is an excellent value considering that CO 2 is being captured in the process. 7. Acknowledgement This work is part of the EU financed project INNOCOUOS (FP7 Contract No. 241401), coordinated by the Chalmers University of Technology (www.clc-innocuous.eu). 8. References [1] Hossain M., de Lasa H. I. Chemical-looping combustion (CLC) for inherent CO 2 separation-a review. Chem. Eng. Sci. 2008;63:4433-4451. [2] Lyngfelt A. Oxygen Carriers for Chemical Looping Combustion - 4 000 h of Operational Experience. Oil & Gas Science and Technology Rev. IFP Energies nouvelles 2011;66(2):161-172. [3] Lyngfelt A., Johansson M., Mattisson T. Chemical-looping combustion Status of development, in: proceedings of the 9th International Conference on Circulating Fluidized Beds 13.-16. May 2008, Hamburg, Germany. [4] Linderholm, C., Mattisson, T., Lyngfelt, A., Long-term integrity testing of spray-dried particles in a 10 kw chemicallooping combustor using natural gas as fuel. Fuel 2009;88:2083-2096 [5] Kolbitsch P., Bolhàr-Nordenkampf J., Pröll T., Hofbauer H. Operating experience with chemical looping combustion in a 120 kw dual circulating fluidized bed (DCFB) unit. Int. J. of Greenhouse Gas Control 2010;4:180-185. [6] Marx, K., Pröll, T., Hofbauer, H. Next Scale Chemical Looping Combustion: Fluidized Bed System Design for Demonstration Unit", in: Proceedings of the 21st International Conference on Fluidized Bed Combustion (FBC), 3.-6. June 2012, Naples, Italy, pp.269-276, ISBN978-88-89677-83-4. [7] Kempkes, V., Kather, A., Chemical looping combustion: Comparative analysis of two different overall process configurations for removing unburnt gaseous components. In: Proceedings of the 2 nd International Conference on Chemical Looping 2012, 26.-28. September 2012, Darmstadt, Germany.