CCPC/CanmetENERGY Phase V

Transcription

1 Final Report of CanmetENERGY Research and Development Prepared for: Canadian Clean Power Coalition (CCPC) Prepared by: Robin Hughes, Research Scientist Marc Duchesne, Research Scientist Robert Symonds, Research Scientist Dennis Lu, Research Scientist February 6, 2017

2 Disclaimer: Neither Natural Resources Canada nor any of its employees makes any warranty express or implied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of its contents. Reference in the report to any specific commercial product, process, service or organization does not necessarily constitute or imply endorsement, recommendation or favouring by Natural Resources Canada. The views and opinions of authors expressed in this report do not necessarily state or reflect those of Natural Resources Canada. Her Majesty the Queen in Right of Canada, as represented by the Minister of Natural Resources, 2017

3 Executive Summary In 2014, Natural Resources Canada CanmetENERGY executed a Task Shared Agreement to provide in-kind R&D associated with the Canadian Clean Power Coalition (CCPC) Phase V work. CanmetENERGY s in-kind contribution included research and development on the topics of chemical and calcium looping, entrained flow gasification, and oxy-pressurized fluidized bed combustion. This report provides a summary of the R&D that CanmetENERGY completed under the Task Shared Agreement. Chemical and calcium looping Calcium looping (CaL) is a process by which CO 2 is separated from flue gas by carbonation of a calcium oxide based sorbent. The carbonated sorbent is regenerated by heating, thus producing a high purity CO 2 stream for utilization or sequestration. Chemical looping combustion (CLC) is a technology for the conversion of fossil fuels with inherent separation of CO 2 and efficient use of energy. Compared to traditional combustion technologies, direct contact between fuel and air is avoided in CLC, as oxygen is transferred from the air reactor to the fuel reactor through an oxygen carrier (typically a metal oxide), and thus pure CO 2 can be obtained without the need for costly gas separation. In an effort to advance CaL and CLC, CanmetENERGY has: Tested CaL sorbent pre-treatments and hydration methods to increase sorbent performance Tested CaL sorbents under realistic operating conditions at the pilot-scale Studied the effect of pressure on the CaL process Determined the performance of sorbents for processes integrating both CaL and CLC Simulated processes combining CaL and CLC Determined the performance of Canadian ilmenite ore under pressurized CLC conditions Entrained flow gasification Entrained flow gasification technology advancement can be enhanced by developing reliable gasifier models and addressing a number of technical hurdles associated with fuel feeding, ash behaviour, and instrumentation. To address some of these issues, CanmetENERGY has completed the following: Developed a comprehensive fluid dynamics model and reactor network models for analysis of gasifier performance Determined physical and thermodynamic properties of liquid and solid ashes Studied pressurized pneumatic conveying of pulverized fuels i

4 Developed and ruggedized instruments for gasifier temperature monitoring Oxy-pressurized fluidized bed combustion CanmetENERGY has focussed its oxy-pressurized fluidized bed combustion (oxy-pfbc) research effort on the following: Developing oxy-pfbc boilers for Oil Sands applications such as steam assisted gravity drainage (SAGD) Developing oxy-pfbc boilers for power generation The oxy-pfbc program will result in the construction and operation of a 1 MW th pressurized pilot scale facility with oxygen firing and test a number of fuels including western Canadian coals and oil sands residues such as petroleum coke. This will advance the technology from TRL 2 to TRL 6. The program will involve comprehensive modeling and validation testing, testing of biomass fuels and blends, testing of system components for suitability in oxy-fuel conditions, enhancing sorbent utilization, and techno-economic assessment of the technology as a prerequisite to commercial demonstration. During Phase V, CanmetENERGY has completed a series of tests at 50 kw th scale to better understand combustion characteristics in oxy-pfbc systems and has nearly completed the construction of the 1 MW th pressurized pilot-scale facility. Application formulated Component prototype demonstration (0.1-5% of full scale) ii

5 Table of Contents EXECUTIVE SUMMARY... I 1 CHEMICAL AND CALCIUM LOOPING CALCIUM LOOPING COMBINED CHEMICAL AND CALCIUM LOOPING PRESSURIZED CHEMICAL LOOPING COMBUSTION ENTRAINED FLOW GASIFICATION INSTRUMENTATION MODELING FUEL FEEDING PHYSICAL AND CHEMICAL PROPERTIES OF ASHES OXY-PRESSURIZED FLUIDIZED BED COMBUSTION MW TH OXY-PFBC PILOT-PLANT FUEL CONVERSION RATE CALCIUM-BASED SORBENT PRESSURIZED SO 2 CAPTURE TECHNOLOGY DEVELOPMENT RISK COMPREHENSIVE FLUID DYNAMICS MODEL PROCESS SIMULATION ASH PARTICLE AGGLOMERATION MODEL ACID DEW POINT MEASUREMENT CHEMICAL & CALCIUM LOOPING PUBLICATIONS EFFECT OF PRESSURE AND GAS CONCENTRATION ON CO 2 AND SO 2 CAPTURE PERFORMANCE OF LIMESTONES POST-COMBUSTION CO 2 CAPTURE BY FORMIC ACID-MODIFIED CAO-BASED SORBENTS PERFORMANCE OF HYDRATION REACTIVATED CA LOOPING SORBENTS IN A PILOT-SCALE, OXY-FIRED DUAL FLUID BED UNIT COMBINED CALCIUM LOOPING AND CHEMICAL LOOPING COMBUSTION CYCLES WITH CAO CUO PELLETS IN A FIXED BED REACTOR PRESSURISED CALCINATION ATMOSPHERIC CARBONATION OF LIMESTONE FOR CYCLIC CO 2 CAPTURE FROM FLUE GASES INVESTIGATING THE USE OF CAO/CUO SORBENTS FOR IN SITU CO 2 CAPTURE IN A BIOMASS GASIFIER CHARACTERIZATION OF AN ILMENITE ORE FOR PRESSURIZED CHEMICAL LOOPING COMBUSTION ATTRITION OF CAO-BASED PELLETS IN A 0.1 MW TH DUAL FLUIDIZED BED PILOT PLANT FOR POST- COMBUSTION CO 2 CAPTURE PRESSURIZED CHEMICAL LOOPING COMBUSTION WITH CO: REDUCTION REACTIVITY AND OXYGEN- TRANSPORT CAPACITY OF ILMENITE ORE CO 2 CAPTURE PERFORMANCE OF CAO BASED PELLETS IN A 0.1 MW TH PILOT-SCALE CALCIUM LOOPING SYSTEM THE EFFECT OF STEAM ADDITION TO THE CALCINER IN A CALCIUM LOOPING PILOT PLANT COMBINED CALCIUM LOOPING AND CHEMICAL LOOPING COMBUSTION FOR POST-COMBUSTION CARBON DIOXIDE CAPTURE: PROCESS SIMULATION AND SENSITIVITY ANALYSIS THE EFFECTS OF THERMAL TREATMENT AND STEAM ADDITION ON INTEGRATED CUO/CAO CHEMICAL LOOPING COMBUSTION FOR CO2 CAPTURE GASIFICATION PUBLICATIONS PARTITIONING OF INORGANIC ELEMENTS IN PILOT-SCALE AND DEMONSTRATION-SCALE ENTRAINED-FLOW GASIFIERS FATE OF INORGANIC MATTER IN ENTRAINED-FLOW SLAGGING GASIFIERS: FUEL CHARACTERIZATION FATE OF INORGANIC MATTER IN ENTRAINED-FLOW SLAGGING GASIFIERS: PILOT PLANT TESTING iii

6 5.4 A SURVEY ON CURRENT STATE-OF-THE-ART IGCC POWER PLANT TECHNOLOGY, SENSORS AND CONTROL SYSTEMS GASIFICATION TEMPERATURE MEASUREMENT WITH FLAME EMISSION SPECTROSCOPY REDUCED ORDER MODELING OF A SHORT-RESIDENCE TIME GASIFIER PARAMETRIC ANALYSIS USING A REACTOR NETWORK MODEL FOR PETROLEUM COKE GASIFICATION THERMODYNAMIC EFFECTS OF CALCIUM AND IRON OXIDES ON CRYSTAL PHASE FORMATION IN SYNTHETIC GASIFIER SLAGS CONTAINING FROM 0 TO 27 WT.% V 2O EXPERIMENTAL ASSESSMENT, MODEL VALIDATION, AND UNCERTAINTY QUANTIFICATION OF A PILOT- SCALE GASIFIER PRESSURIZED PNEUMATIC CONVEYING OF PULVERIZED FUELS FOR ENTRAINED FLOW GASIFICATION PETROLEUM COKE GASIFICATION TEMPERATURES AND FLAME SPECTRA IN THE VISIBLE REGION AT HIGH PRESSURE SLAG DENSITY AND SURFACE TENSION MEASUREMENTS BY THE CONSTRAINED SESSILE DROP METHOD OXY-FUEL FLUIDIZED BED CONVERSION PUBLICATIONS COMBUSTION CHARACTERISTICS OF COAL AND COKE UNDER HIGH OXYGEN CONCENTRATION OXY-FUEL CFBC CONDITIONS OXY-FLUIDIZED BED COMBUSTION USING UNDER BED FINES FUEL INJECTION HIGH TEMPERATURE MONITORING OF AN OXY-FUEL FLUIDIZED BED COMBUSTOR USING FEMTOSECOND INFRARED LASER WRITTEN FIBER BRAGG GRATINGS EFFECTS OF CO 2 ON THE FUEL NITROGEN CONVERSION DURING COAL RAPID PYROLYSIS iv

7 Tables and Figures Table 1. Base case - principal results... 7 Table 2. Linear combination fitting results for gasifier and synthetic samples Figure 1. Schematic of CanmetENERGY's CaL pilot plant... 2 Figure 2. Effect of total pressure on (a) CO 2 capture and (b) SO 2 capture of Florina limestone... 3 Figure 3. CO2 carrying capacity of raw, spent and reactivated sorbents Figure 4. Process arrangements for combined chemical and calcium looping technologies Figure 5. The 15th carbonation/oxidation/calcination/reduction cycle of Cu50-Ca40- Cem10 pellets followed by a subsequent oxidation step Figure 6. Micrographs of CaO/CuO pellets after (a) 0; (b) 5; (c) 10; and (d) 15 cycles of CaL and CLC. Calcium (Ca) and copper (Cu) are represented in purple and yellow, respectively Figure 7. Migration of iron to the surface of ilmenite ore after calcination; (a) raw and (b) calcined Figure 8. Effects of total pressure and temperature on ilmenite ore's conversion rate. 10 Figure 9. Photograph of the high pressure LIBS cell Figure 10. Reactor temperature as a function of position and time, during part of one thermal cycle, plotted with 10 minute resolution. Data from fibers 1-3 with 2 cm sensor spacing Figure 11. CFD simulation vector plots (a) Case 1 (b) Case 2 (c) Case Figure 12. Sensitivity analysis using ROM of (a) carbon conversion; (b) of H2/CO ratio; (c) outlet temperature; (d) peak temperature Figure 13. Dry syngas compositions of the experimental tests compared to ROM simulation results Figure 14. Configuration of feed hopper in the cold-flow dense-phase conveying system at Figure 15. (a) Conventional and (b) constrained sessile drop contours Figure 16. Surface tension of synthetic coal slag measured on 25, 13 and 8 mm substrates of molybdenum in Ar/H Figure dimensional model of the oxy-pfbc plant with photographs of core components Figure 18. Predicted cost of electricity....error! Bookmark not defined. Figure 19. Particle residence time at 9 and 12 bar pressure v

8 Figure 20. Pressurized fluidized bed used for residence time measurements Figure 21. A snapshot of a fluid bed CFD transient model with blue indicating regions that are predominantly solids Figure 22. Solidus temperatures of possible eutectics found in coal ash Figure 23. Predicted liquid fraction versus temperature for agglomeration samples from the Canmet 50 kwth oxy-fbc vi

9 1 Chemical and Calcium Looping 1.1 Calcium looping Calcium looping in dual fluidized beds was proposed by Tadaaki Shimizu from Niigata University in Japan in 1999 as a post-combustion CO2 capture technology that was intended to be a higher efficiency and lower cost replacement for the suite of amine technologies. Shortly afterwards, technology development began at CanmetENERGY with pilot plant testing occurring in a single fluidized led by Ben Anthony (CanmetENERGY) and Carlos Abanades (Spanish National Research Council). Initially the technology was to be applied to coal fired power plants as a means of capturing 90+% of CO2 produced by the power plant. By 2005 CanmetENERGY demonstrated the use of the technology at the 50 kw th thermal scale and was able to routinely demonstrate high CO 2 capture rates (96%) in a dual fluid bed system with an oxygen fired-calciner (similar to the arrangement seen in Figure 1). This test work demonstrated that the major technical challenges associated with the technology included CO 2 capture sorbent attrition, sorbent poisoning by sulphur and sintering. During this period the technology became increasingly important to CCS researchers internationally. In parallel to this, techno-economic evaluations were completed which indicated that the cost of CO 2 capture was attractive in comparison with competing post combustion CO 2 capture technologies. Subsequently, larger scale demonstrations of the technology were completed in Germany (TU Darmstadt, 1MW th; University of Stuttgart 0.2 MW th), Spain (La Pereda 2013, 1.7 MW th), and Taiwan (ITRI 1.9 MW th). It is believed that the economics of calcium looping technology are most attractive when the technology is built in conjunction with a cement plant which uses the spent sorbent (i.e. lime) from the CO 2 capture plant as the primary feedstock to the cement plant which simultaneously decarbonizes the power plant and greatly reduces the CO 2 emissions associated with the cement plant. The calcium looping R&D described in this report is intended to address the technical challenges addressed above; sintering, effect of sulphur and attrition. Interest in calcium looping technology has increased for the purpose of capturing CO2 from natural gas fired facilities, so CanmetENERGY has completed related work as part of the CCPC Phase V effort which is reported in CCPC s report on Evaluation of Options for Natural Gas Fired Power Plant with Carbon Capture. 1

10 1.1.1 Pilot-scale testing and sorbent pre-treatment Pilot-scale calcium looping experiments were completed to compare the performance of different limestone-based pellets to unmodified limestones with continuous sorbent make-up to represent realistic operating conditions (Figure 1). The lower than expected performance of limestone-based pellets was attributed to the rapid and severe sintering of the pellets at the early stages of the test and significant particle attrition. It was concluded that, under the conditions tested, the use of limestone-based pellets does not seem to be a promising approach due to the significantly higher cost of production with comparable CO 2 capture performance to that of unmodified limestone. Furthermore, the attrition of pellets was found to be comparable to that of limestone tested under similar conditions, with the attrition extent being relatively insensitive to the limestone type used for pellets or the sorbent injection method used. The attrition trends suggest that the improvement in the mechanical strength of the pellets was marginal compared to that of unmodified limestone. Figure 1: Schematic of CanmetENERGY's CaL pilot-plant. The performance of limestone-based sorbents modified with formic acid was also investigated. In the first calcium looping cycle, limestone treated with a 10% acid solution displayed a CO2 capture capacity of 0.6 g CO2/g sorbent, compared to 0.49 g/g for untreated limestone. After twenty cycles, the modified sorbent maintained a higher carrying capacity, capturing approximately 67% more CO2 than the untreated limestone. 2

11 1.1.2 Effect of pressure The performance of two Greek limestones were evaluated by performing looping cycles via atmospheric and pressurized thermogravimetric analysis (TGA) under varying partial pressures of CO2 and SO2 to simulate combustion flue gas. Increasing the pressure from atmospheric to 10 bar led to deterioration in CO2 capture performance (Figure 2a). Further pressure increment had negligible effect on the CO 2 capture performance of the limestones. SO 2 retention, however, improved with increasing pressure (Figure 2b). A separate study looked at the effect of increasing the total pressure for calcination, while maintaining atmospheric pressure for carbonation. The results indicate that the carbonation conversion of calcined sorbent decreases as the pressure is increased during calcination. Pressurised calcination requires higher temperatures and causes an increase in sorbent sintering, albeit that it would have the advantage of reducing equipment size as well as the compression energy necessary for CO 2 transport and storage. (a) Figure 2: Effect of total pressure on (a) CO 2 capture and (b) SO 2 capture of Florina limestone Sorbent hydration (b) The progressive deactivation of limestone-based sorbents is one of the major limitations of the calcium looping cycle. Hydration of spent sorbent has been identified as a promising route for their reactivation (Figure 3). The performance of sorbents reactivated by two different hydration techniques was assessed in CanmetENERGY s CaL pilot-plant. Compared to a spent sorbent, reactivated materials exhibited a ~60% increase in CO 2 carrying capacity over 3 h of circulation, as well as an increase in rate of particle attrition. It was noted that this increased attrition did not lead to any serious disruptions in system operation. 3

12 Figure 3: CO2 carrying capacity of raw, spent, and reactivated sorbents. 1.2 Combined chemical and calcium looping Techno-economic analyses have indicated that the largest parasitic power consumer and highest capital cost portion of calcium looping post combustion CO 2 capture facilities is the air separation unit required for the oxy-fired calciner which regenerates the CO2 capture sorbent. For this reason a variety of methods of providing the heat required to calcine the sorbent have been considered. In this report we present R&D that has been performed in which this heat is provided through chemical looping combustion (CLC). CLC could replace oxy-fuel combustion for CO 2 sorbent regeneration via the use of a metal oxide (such as copper(ii)-oxide, CuO) acting as an oxygen carrier. Process arrangements for combined CLC and CaL applied to post combustion CO 2 capture, sorption enhanced reforming, and gasification are provided in Figure 4. 4

13 Figure 4: Process arrangements for combined chemical and calcium looping technologies. Bench-scale experiments were conducted to investigate the integration of CaL and CLC with steam gasification of biomass. It was demonstrated that the use of composite CaO/CuO/calcium-aluminate-cement pellets for gasification purposes required oxidation of Cu to be preceded by carbonation as opposed to the post-combustion case where the pellets are oxidized prior to carbonation. Composite pellets were, thus, tested under this CO 2 capture sequence using varying carbonation conditions over multiple cycles (Figure 5). While the pellets exhibited relatively high carbonation conversion, the oxidation conversion declined for all tested conditions likely due to the CaCO 3 product impeding passage of O2 molecules to the more remote Cu sites. The reduction in oxygen uptake was particularly important when the pellets were pre-carbonated in the presence of steam. Limestone-based pellets and Cu-based pellets were subsequently tested in separate CaL and CLC loops respectively to assess their performance in a dual-loop process. 5

14 Figure 5: The 15th carbonation/oxidation/calcination/reduction cycle of Cu50-Ca40Cem10 pellets followed by a subsequent oxidation step. The deactivation behavior of CaO, CuO, and integrated CuO/CaO pellets was studied (Figure 6). With cycling, copper migrated to the surface of the composite pellets, which likely suppressed carbonation capacity by reducing the accessibility of the CaO. While thermal pre-treatment and steam addition enhanced the performance of the base CaO pellets, the former led to cracks in the pellets. In contrast, thermal pre-treatment of the CuO/CaO composite pellets resulted in reduced CLC and CaL performance. Figure 6: Micrographs of CaO/CuO pellets after (a) 0; (b) 5; (c) 10; and (d) 15 cycles of CaL and CLC. Calcium (Ca) and copper (Cu) are represented in purple and yellow, respectively. 6

15 A CaL-CLC process was simulated and compared, in terms of efficiency, power production and solids circulation rates, to a process using CaL alone. In addition, a new dual loop configuration for solids looping in the CaL-CLC process was proposed with the purpose of mitigating the loss of calcium oxide capacity with cycling. Simulations showed an improved process efficiency of the CaL-CLC compared to CaL alone and an increased power output due to the higher solids circulation rates (Table 1). The dual loop configuration provides the same higher efficiency, but with a reduced solids flow rate compared to the conventional CaL-CLC configuration. A sensitivity analysis of the process operating parameters was performed and partial CO 2 capture (down to 60%) scenarios were considered. Table 1: Base case Principal results. Fuel input in calciner (LHV, MWth) Extra power (MWe) Penalty (LHV%) Solids flow rate (kg/s) CaL-CLC single loop % 2354 CaL-CLC dual loop % 2122 CaL % 143 While the efficiency penalties of combined calcium and chemical looping look promising, the required solids flow rates are very high. These high flow rates in combination with the complexity of operating three integrated fluidized beds make it unlikely that the combined calcium and chemical looping technology will be commercialized. 1.3 Pressurized chemical looping combustion Chemical looping combustion is a means of burning fuels in which air does not come into direct contact with the fuel. Instead, an oxygen carrier is oxidized in an air reactor and is then transferred to the fuel reactor where the oxygen reacts with the fuel. The gas leaving the air reactor is enriched in nitrogen whereas the gas leaving the fuel reactor is composed entirely of the products of combustion of the fuel. The process is equivalent to oxy-combustion of fuels without the need for a costly air separation unit. The main technical challenges in chemical looping combustion are oxygen carrier attrition, poisoning of the oxygen carrier by ash components and sulphur, and achieving sufficiently high fuel conversion rates. In order to address these issues CanmetENERGY has been investigating the potential of pressurized chemical looping combustion using ilmenite as an oxygen carrier for various applications. Ilmenite has been selected in part since it has been shown that it is not negatively affected by ash components and sulphur, and so it is suitable for a wide array of fossil fuels. 7

16 CanmetENERGY is currently working on the following topics related to pressurized chemical looping combustion: Development of mesoporous coatings to improve oxygen carrier performance. Early results indicate that adding 1-10 wt-% of manganese, cobalt, nickel or copper to the ilmenite ore improves the conversion rate of methane. Oxygen carrier oxidation reactivity testing and kinetic / reactor modeling Comprehensive computation fluid dynamics (CFD) modeling of both the fuel and air reactor. Optimization of process configuration for steam assisted gravity drainage (SAGD) via comprehensive process simulation and heat integration Basic and detailed engineering towards the design, construction, commissioning, and operation of a 600 kwth pilot-plant located at CanmetENERGY Reduction reactivity testing The reduction reactivity and oxygen-transport capacity of Canadian ilmenite ore used for pressurized chemical looping combustion were investigated in a pressurized thermogravimetric analyzer (PTGA). The first series of tests used a CO-CO 2-N2 mixture for reduction, and air for oxidation. Increasing the temperature from 1123 to 1323 K resulted in a pronounced increase in the reduction rate. Although increasing the particle size from micron to micron resulted in a slower rate of reduction, the effect was minor. X-ray diffraction (XRD) comparing raw and calcined ore samples indicate the disappearance of ilmenite crystals and the formation of rutile and ferric pseudobrookite. Scanning electron microscopy (SEM) images reveal the development of cracks in the oxidized particles after 4 redox cycles at 950 C and 16 bar, which was correlated to the original lamellar structure of the raw ilmenite ore (Figure 7). Total pressure did not have a noticeable effect on the surface morphology and fuel partial pressure did not result in any significant changes on the surface morphology. Increasing the number of redox cycles from 3.5 to 19.5 resulted in large cracks near iron-rich lamellae. Increasing the reaction temperature from 850 C to 1050 C resulted in similar surface morphologies, but larger grains were formed with increasing reaction temperature. 8

17 Figure 7: Migration of iron to the surface of ilmenite ore after calcination; (a) raw and (b) calcined. Reduction kinetics of Canadian ilmenite ore as an oxygen carrier for methane chemical looping combustion under elevated pressure was studied using pressurized thermogravimetric analysis (PTGA). The reduction phase of the experiments was carried out in a mixture of methane and nitrogen with carbon dioxide and/or steam to simulate an actual combustion environment. The oxidation phase of the experiments was carried out with air. Effects of temperature ( K), total pressure (6 16 bar), methane partial pressure ( bar), and particle size ( µm) were studied (Figure 8). Tests were carried out to examine the effect of higher redox cycle numbers on the performance of ilmenite ore. The results showed that increasing the total pressure reduced the rate of conversion during ilmenite ore reduction. A kinetic model, based on a phase-boundary controlled mechanism with contracting sphere was developed in the Arrhenius form using these experimental data. 9

18 Figure 8: Effects of total pressure and temperature on ilmenite ore's conversion rate. 10

19 2 Entrained Flow Gasification Entrained flow gasification for power production is commercially available technology; however, the technology appears to be quite expensive in its currently commercial form. CanmetENERGY s gasification R&D activities discussed in this report are intended to reduce the cost of entrained flow gasification through the advancement of lower capital cost and higher efficiency gasifiers, advanced instrumentation, and reactor network modeling. Furthermore, CanmetENERGY has continued R&D to determine parameters required for high quality design work of gasifiers firing Canadian fuels. 2.1 Instrumentation Spectroscopy techniques High pressure gasification of fossil fuels produces syngas. Syngas is mainly composed of hydrogen, carbon monoxide, carbon dioxide, and nitrogen with small amounts of other gases or trace elements. In a gasification plant, the syngas exiting the gasifier is cleaned before subsequent use. The laser induced breakdown spectroscopy (LIBS) technique can be used to test the syngas before and after cleaning and detect the presence of any unwanted elements in the gas. A high pressure (max. 15 bar) gas cell was designed and manufactured to perform LIBS measurements at various pressures (Figure 9). In our tests, a plasma is generated by a high power pulsed laser. Our results indicate that hydrogen, nitrogen, oxygen and carbon can be detected with the LIBS technique at high pressures in situ, and LIBS may be used to estimate relative concentrations of the four elements. In addition to LIBS work, we also completed flame emission spectroscopy (FES) and tunable diode laser absorption spectroscopy (TDLAS) work. 11

20 Figure 9: Photograph of the high pressure LIBS cell Fiber Bragg grating arrays Femtosecond pulse duration infrared laser (fs-ir) written fiber Bragg gratings (FBGs), have demonstrated great potential for extreme environment sensing. Harsh environments are inherent to the advanced power plant technologies under development to reduce greenhouse gas emissions. The performance of new power systems are currently limited by the lack of sensors and controls capable of withstanding the high temperature, pressure, and corrosive conditions present during gasification or combustion. We fabricated and deployed several fs-ir written FBG arrays for monitoring the temperature distribution within a reactor. Results include: calibration data to 1100 C, discussion of deployment strategies, contrast with thermocouple data, and comments on reliability (Figure 10). 12

21 Figure 10: Reactor temperature as a function of position and time, during part of one thermal cycle, plotted with 10 minute resolution. Data from fibers 1-3 with 2 cm sensor spacing. 2.2 Modeling Computational fluid dynamics (CFD) model simulations High-pressure entrained flow gasifier technology is used to convert solid carbonaceous feedstocks into synthesis gas, which can be used in an integrated gasification combined cycle power plant or as a feedstock for chemical or synthetic fuel production. Computational fluid dynamics (CFD) models, once validated, can be used to help design full-scale reactors. Model validation entails the comparison of model predictions to labscale or pilot-scale measurements. However, experimental measurements of highpressure gasifiers usually consist only of wall temperatures and outlet gas temperature and composition, which are of limited use for model validation when the gasifier is operating well, providing information only about operating temperature, heat loss, and equilibrium gas composition. These do not provide a strong validation of the CFD model, whose main purpose is to make predictions of the flame size and shape, and its ability to convert solid fuel to gas efficiently in a small volume. We performed model validation based on data generated using CanmetENERGY s high-pressure entrained flow gasifier. To provide a stronger validation, the approach taken was to compare the model predictions over a range of operating conditions comprising both favorable and unfavorable performance. The CFD model is able to track the performance of the gasifier over the range of operating conditions; it is able to predict higher and lower solid fuel conversion levels depending on the operating conditions, distinguishing favorable versus unfavorable operating conditions and additionally providing insight into the causes (Figure 11). 13

22 Figure 11: CFD simulation vector plots (a) Case 1 (b) Case 2 (c) Case Reduced order model (ROM) simulations Reduced order models (ROMs) and reactor networks are becoming widely accepted tools for the modeling of complex reactors such as entrained-flow gasifiers. The approximations made in a ROM reduce the required computational costs compared to Computational Fluid Dynamic (CFD) models; however; the reliability of the model in predicting the outputs for a range of operating conditions in the gasification unit faces challenges. We compared simulation results between a ROM and the corresponding CFD model of a short-residence time gasifier under different operating conditions and kinetic parameters. Although the proposed ROM s framework was fixed and developed based on CFD simulations generated at a base-case condition, the results showed reasonable agreement between the two models in predicting syngas composition, carbon conversion and the temperature profile in the gasification system. Sensitivity analysis of the ROM s inputs (including test condition and reactor network parameters) has also been performed. This analysis has shown that the recirculation ratio and oxygen flowrate have a greater effect on the outputs compared to model geometry and kinetic parameters. Dry gas composition and flowrate, temperature distribution, conversion and pollutant formation taken from the experimental tests were used to validate the prediction capabilities of the ROM (Figure 13). The ROM predicted the experimental observations for conversion in the range of 48-90%. The ROM can predict the behaviour of a gasifier under different operating conditions with reasonable accuracy. Moreover, we 14

23 investigated the variability in the ROM s key outputs in the presence of uncertainty in the feed and model parameters, i.e., the volatile percentage of the fuel, solid particle diameters, angle of multi-phase flow jet and recirculation ratio (Figure 12). These parameters affect the feedstock s properties and the mixing/laminar flows within different zones of the gasifier. Insights gained from the uncertainty quantification study revealed significant variability in the conversion, peak temperature and steam percentage in the syngas; while the dry syngas composition seems to not be significantly affected by the uncertainty of the parameters considered. Figure 12: Sensitivity analysis using ROM of (a) carbon conversion; (b) of H2/CO ratio; (c) outlet temperature; (d) peak temperature. 15

24 Figure 13: Dry syngas compositions of the experimental tests compared to ROM simulation results. 2.3 Fuel feeding We designed, built, and tested a pneumatic conveying system for the continuous conveying of pulverized fuels, namely biomass (in the form of Canadian forestry waste), lignite, and petroleum coke, to an entrained flow gasification unit (Figure 14). All three fuels studied exhibit properties that are challenging for fluidization or conveying. The lignite and petroleum coke are Geldart class C particles and are expected to exhibit difficulties in conveying, while the biomass is a Geldart class A particle but the nonuniform particle shape and large size distribution result in pneumatic conveying challenges similar to class C particles. The conveying system consists of a blow vessel with three points of gas injection to aerate the bed material and facilitate hopper discharge. The mass flux of each material was studied as the system parameters were varied, which included: fluidizing and sparge gas (two of the three gases aerating the hopper bed), transfer gas (gas injected directly into the solid transfer line), pressure drop, conveying gas type and transfer line diameter. Fuel fluxes were varied in a range of 450 kg/m2s to 1700 kg/m2s for all three fuels. Compared to lignite and petroleum coke conveying, biomass conveying was found to have a smaller gain in mass flux to many conveying parameters investigated. Several models for pneumatic conveying of powders were compared against the data obtained. The models were found to have various degrees of relative error, with the best fitting model having a relative error of less than 10% for all three fuels with a marginal bias towards underestimation. 16

25 Figure 14: Configuration of feed hopper in the cold-flow dense-phase conveying system at CanmetENERGY. 2.4 Physical and chemical properties of ashes Enhanced slag density and surface tension measurement method Physical properties of slag are critical in the design and operation of refining technologies and slagging energy systems. The density and surface tension of slag impacts phenomena such as granulation, foaming, removal of solid inclusions, erosion of refractory, and fouling. We compared slag sessile drops formed on graphite, alumina and molybdenum substrates (Figure 15). Use of graphite resulted in the largest contact angles, a desirable trait for density and surface tension measurements, but also led to reactions with the slag. Alumina and molybdenum were less reactive, but resulted in contact angles too small for measurements. When sessile drops were constrained by small substrate diameters to increase the apparent contact angle, surface tension and density measurements could be achieved with alumina and molybdenum substrates. The surface 17

26 tension of coal slag was measured at up to 1600 C in oxidizing and reducing gas atmospheres (Figure 16). (a) (b) Figure 15: (a) Conventional and (b) constrained sessile drop contours. Figure 16: Surface tension of synthetic coal slag measured on 25, 13 and 8 mm substrates of molybdenum in Ar/H Obtained and analysed X-ray absorption data for gasification samples with vanadium Vanadium is found in slags produced during metal refinement and fossil fuel combustion/gasification. The oxidation state of vanadium in slag has technological and environmental implications. For example, it may affect slag flow and refractory wear inside reactors, as well as leachability and toxicity of industrial by-products. Determination of vanadium s oxidation state in crystalline phases can be achieved via the widely adopted X-ray diffraction (XRD) technique. However, this technique does not provide information on vanadium in amorphous phases. The objective of this research is to determine the oxidation state of vanadium in petroleum coke gasification samples and laboratory samples using X-ray absorption spectroscopy (XAS) with Canadian Light 18

27 Source s soft X-ray micro-characterization beamline (SXRMB). Linear combination fitting of XAS spectra with reference samples allowed quantitative determination of vanadium speciation (Table 2). Table 2: Linear combination fitting results for gasifier and synthetic samples. 19

28 3 Oxy-Pressurized Fluidized Bed Combustion CanmetENERGY began development of oxy-fluidized bed combustion (oxy-fbc) in 2003 as an alternative to pulverized fuel oxy-combustion technology. It was believed that oxyfbc had the potential to be more cost competitive than pulverized fuel oxy-combustion due to the ability of fluidized bed combustors to remove heat from the combustion zone more rapidly. This would allow higher oxygen concentrations to be used in the oxidant supply and hence would require lower flue gas recycle rates. This would result in much smaller and hence lower cost power plants. By 2004, CanmetENERGY demonstrated the technology at the 50 kwth scale in both bubbling bed and circulating bed modes with recycled flue gas. In 2005, CanmetENERGY successfully operated the oxy-fbc with oxygen concentrations as high as 55 vol% proving that compact equipment design was possible. In 2008, CanmetENERGY completed the construction of a 0.8 MWth oxy-cfbc and operated the facility in collaboration with Foster Wheeler to generate design data required for a demonstration plant. The demonstration plant was built by Foster Wheeler and successfully operated in Spain starting in Foster Wheeler is now prepared to provide the technology at 300+ MWe scale. In parallel to these activities, CanmetENERGY recognized that there was the potential to further reduce the cost of oxy-fbc technology by increasing the efficiency of the technology through pressurization with the first concept diagram being completed in In 2010 CanmetENERGY began planning and design work for an oxy-pressurized fluidized bed combustion (oxy-pfbc) pilot plant performing multiple process performance analyses. In 2012, Dr. Duke Duplessis encouraged CanmetENERGY to collaborate with Aerojet Rocketdyne on development of oxy-pfbc technology, which was also developing oxy-pfbc technology with the intent to construct a pilot plant of similar scale. CanmetENERGY, Aerojet Rocketdyne and Linde began collaboration on pilot plant construction in Linde is responsible for development of the CO2 purification section of the facility which includes latent heat recovery, SOX & NOX removal and deoxidation. Aerojet Rocketdyne s oxy-pfbc technology and team have now been transferred to the Gas Technology Institute and the co-development of the technology has continued. The R&D activities discussed in this report summarize CanmetENERGY s recent oxy-pfbc development activities with an emphasis on ensuring the technology is suitable for power generation with Canadian coals and biomass. 20

29 3.1 1 MWth oxy-pfbc pilot-plant The construction of the 1 MWth oxy-pressurized fluidized bed combustion pilot plant is complete and the facility has been commissioned (Figure 17). Combustion tests firing Illinois #6, Poplar River, and Genesee coals will proceed throughout Figure 17: Process flow diagram of the 1 MWth oxy-pfbc pilot plant at CanmetENERGY 3.2 Fuel conversion rate Collaboration was established with the University of Ottawa in which a Master s student worked with a pressurized fluidized bed (Figure 19) to determine the elutriation rates of fuel particles under various conditions and subsequently the residence times of the particles (Figure 18). The facility used to perform this test work was modified with the help of CanmetENERGY to inject fine particles. 21

30 Figure 18: Particle residence time at 9 and 12 bar pressure. 22

31 Figure 19: Pressurized fluidized bed used for residence time measurements. 3.3 Calcium-based sorbent pressurized SO2 capture More than 50 sulphation tests have been completed using a pressurized thermogravimetric analyzer (PTGA) at CanmetENERGY. The tests were conducted to examine the effect of total pressure, temperature, sorbent type (limestone, dolomite, 23

32 etc.), particle size, and SO2 partial pressure on the reactivity and kinetics of sulphation under oxy-pfbc conditions. A sulphation model has been selected that accounts for initial sorbent particle morphology (surface area and pore volume). Reactivity and kinetic analysis is on-going and is expected to be completed in Technology development risk A series of tests have been completed using Canadian coals in CanmetENERGY s 50 kwth oxy-fbc with 78 test conditions completed. The facility met all performance expectations with fuel conversion up to 98%, no agglomeration during steady state operations, and SO 2 + SO3 concentrations in the flue gas within allowable limits to meet materials constraints. A provisional patent has been issued to CanmetENERGY for a System and Method for Oxygen Carrier Assisted Oxy-Pressured Fluidized Bed Combustion. The patent describes a method that utilizes an oxygen carrier, such as is typically found in chemical looping combustion of solid fuels, to distribute oxygen throughout the fluid bed in order to reduce or eliminate the presence of localized reducing conditions. This reduces risks and costs associated with fuel distribution, corrosion, erosion, explosions, emissions, oxygen consumption, and CO2 purification. The method has been proven to be effective through pilot-plant tests in CanmetENERGY s 50 kwth oxy-fbc. 3.5 Comprehensive fluid dynamics model Transient two-dimensional comprehensive fluid dynamics (CFD) simulations have been completed representing the lower section of the 1MWth oxy-pfbc pilot-plant. Five separate variants of the system have been modeled to better understand what the possible operating issues may be in the pilot-plant and to understand what simplifying assumptions can be made to the model to reduce the computational effort required. Thus far, simulations have been performed to better understand hydrodynamic behaviour within the system via non-reacting simulations (Figure 20). The model indicates that adjustments may be necessary to the combustor geometry to improve heat and mass transfer within the system. The models will be validated during pilot-plant commissioning. 24

33 Figure 20: A snapshot of a fluid bed CFD transient model with blue indicating regions that are predominantly solids. 3.6 Process simulation Process simulations of oxy-pfbc have been completed for pilot and demonstration scales firing five Canadian solids feedstocks and Illinois #6. The results have been used to design the 1 MWth pilot-plant. A steam Rankine cycle process simulation has been completed and will be integrated with the oxy-pfbc simulation. Further studies will examine the impact of start-up and show-down regimes on system performance via transient simulations 3.7 Ash particle agglomeration model The initial particle agglomeration research effort was directed towards establishing the temperatures at which liquids may form in coal ashes of interest. We have considered 25

34 binary and ternary (two- and three-component) systems using thermodynamic software to predict solidus temperatures (Figure 21). Figure 21. Solidus temperatures of possible eutectics found in coal ash. Based on the thermodynamic calculations we have selected eutectics of concern for which we have initiated experimental work to verify that the components do in fact melt at the predicted temperatures and to determine physical parameters of importance to agglomeration modeling including surface tension and viscosity. It was expected that during 50 kwth oxy-fbc tests that were performed that bed agglomeration would occur at conditions when both the oxygen concentration and operating temperature were high (Figure 22). We did not however experience bed agglomeration except once during a pilot-plant upset. The agglomerate was analyzed and a portion shipped to Pennsylvania State University for further analysis. The pilot-plant was operated for extended periods of time at about 900 C without agglomeration, so further work is needed in understanding why agglomeration only occurred once in several hundred hours of pilot-plant operations. 26

35 Figure 22: Predicted liquid fraction versus temperature for agglomeration samples from the Canmet 50 kwth oxy-fbc. 3.8 Acid dew point measurement A literature review was completed on SO3 monitoring technologies for combustion applications to establish what the best method would be to determine SO3 concentration and acid dew point in the oxy-pfbc. An acid dew point sensor (Lancom 200) was purchased and tested in two separate atmospheric pressure pilot facilities. Tests with the 50 kwth oxy-fbc were not successful in measuring an acid dew point temperature using the Lancom 200 sensor; however, the controlled condensation method was successfully used to determine the SO3 concentration (1 to 5.2 ppm) in the flue gas. Further tests applying the Lancom 200 were performed with a pulverized coal combustor with 1) SO 2 injected into the combustor fired with natural gas and 2) fired with coal. During these tests, the Lancom 200 was only able to provide a single acid dew point temperature. It is not clear at this time whether the Lancom 200 will be suitable for use in the oxy-pfbc pilot plant. Obtaining SO3 measurements and acid dew point temperatures is very important to the development of the technology; therefore, we are considering alternate sensors. 27

36 4 Chemical & Calcium Looping Publications 4.1 Effect of pressure and gas concentration on CO2 and SO2 capture performance of limestones Basinas P, Wu Y, Grammelis P, Anthony EJ, Grace JR, Jim Lim C. Effect of pressure and gas concentration on CO2 and SO2 capture performance of limestones. Fuel 2014;122: doi: /j.fuel Two Greek limestones with different properties were examined, to determine their CO2 and SO2 capture performance. The reversibility of the sorbents for CO2 capture was investigated by performing looping cycles in atmospheric and pressurized thermogravimetric reactors, with synthetic gas mixtures containing different partial pressures of CO2 and SO2 to simulate flue gases. The morphological and porosity characteristics of the original and spent sorbent are examined by Scanning Electron Microscopy and Pore Sized Distribution analyses. Increasing pressure from atmospheric to 10 bar led to deterioration in CO2 capture performance. Further pressure increment had negligible effect on the CO 2 capture performance of the limestones. SO2 retention, however, improved with increasing pressure. For calcium looping involving repeated cycles, sorbents sulphated via the unreacted-core mode converted more available calcium, but this adversely affects the reversibility of the cyclic CO2 capture. The reversibility strongly deteriorated when the higher total pressure was combined with increased SO 2 partial pressure. The CO2 uptake of an unreacted-core sulphated sorbent, previously used for SO2 retention, was mainly affected, apart from pore blockage or sintering, by the occupation of calcium. Sulphation under simultaneous capture resulted in higher CO2 removal efficiency for uniformly and network sulphated particles compared to the sulphur capture via direct sulphation. 4.2 Post-combustion CO2 capture by formic acidmodified CaO-based sorbents Ridha FN, Manovic V, Wu Y, Macchi A, Anthony EJ. Post-combustion CO2 capture by formic acid-modified CaO-based sorbents. International Journal of Greenhouse Gas Control 2013;16:21 8. doi: /j.ijggc The performance of CaO-based sorbents modified with formic acid in both its liquid and vapor phase has been investigated for high-temperature post-combustion CO2 capture in calcium-looping cycles. The treatment of limestone with aqueous solutions containing 10 or 30 vol% formic acid was found to promote crystal growth. By contrast, higher acid 28

37 concentrations produced smaller crystals. However, all sorbents modified by acid solutions had almost identical reductions of 44% and 46% in surface area and pore volume (determined by N2 adsorption), respectively, relative to the parent material. Despite the low porosity, limestone (fine powder) treated with 10% acid solution displayed the highest CO2 capture capacity in the first cycle with a capture of 0.6 g CO 2/g sorbent compared to 0.49 g/g for untreated powder material. By 20 cycles, the modified sorbent still captured 67.4% more CO2 than the natural sorbent captured under similar conditions. Relatively low concentration formic acid solution improved the CO 2 capture capacity of CaO-based sorbents better than treatment with acid vapor due to the limited acidification achieved by vapor phase treatment. 4.3 Performance of hydration reactivated Ca looping sorbents in a pilot-scale, oxy-fired dual fluid bed unit Materić V, Symonds R, Lu D, Holt R, Manović V. Performance of Hydration Reactivated Ca Looping Sorbents in a Pilot-Scale, Oxy-fired Dual Fluid Bed Unit. Energy Fuels 2014;28: doi: /ef501203v. The progressive deactivation of CaO based sorbents is one of the major limitations of the Ca Looping cycle for post-combustion CO2 capture. Techniques using the hydration of the spent CaO sorbent have been identified as a promising route for the reactivation of CaO sorbents but have so far not been tested at pilot scale using realistic calcination and carbonation conditions. In this work, the performance of sorbents reactivated by two different reactivation techniques (hydration dehydration and superheating) was assessed in an oxy-fired, pilot-scale dual fluid bed unit. Compared to a spent sorbent, reactivated materials exhibited a 60% increase in CO2 carrying capacity over 3 h of circulation as well as an increase in the sorbent attrition rate of 25% (superheating) and 50% (hydration dehydration). In both cases, however, increased attrition did not lead to serious disruptions of system operation. A comparison of sorbent performance at laboratory and pilot scale suggested that high velocity impacts in the transfer lines were the main cause of attrition. 4.4 Combined calcium looping and chemical looping combustion cycles with CaO CuO pellets in a fixed bed reactor Ridha FN, Lu D, Macchi A, Hughes RW. Combined calcium looping and chemical looping combustion cycles with CaO CuO pellets in a fixed bed reactor. Fuel 2015;153: doi: /j.fuel The feasibility of using integrated CaO/CuO-based pellets in combined calcium and chemical looping cycles in a fixed bed was investigated with emphasis on CO2 capture performance. Three types of pellets were tested, namely; integrated core-in-shell CaO/CuO-based pellets; integrated homogeneous CaO/CuO-based pellets; and mixed 29