FEDERAL AGENCY TO WHICH THE REPORT IS SUBMITTED U.S. Dept of Energy/National Energy Technology Laboratory (DOE/NETL) FEDERAL GRANT NUMBER DE-FE00031722 PROJECT TITLE Transformational Sorbent-Based Process for a Substantial Reduction in the Cost of CO2 Capture PRINCIPAL INVESTIGATOR Ravi Jain, President Telephone: (908) 450-9840 E-mail: ravi.jain@innosepra.com DATE SUBMITTED December 2024 DUNS NUMBER 808159011 RECIPIENT ORGANIZATION InnoSepra, LLC 452 Lincoln Blvd Middlesex, NJ 08846 GRANT PERIOD May 1, 2019 - August 31, 2024 REPORTING PERIOD END DATE August 31, 2024 REPORTING TERM Final Report SIGNATURE OF THE SUBMITTING OFFICIAL 1 TABLE OF CONTENTS Disclaimer 7 Summary 8 Project Accomplishments 11 Task 1: Project Management and Planning 11 Task 2: Technology Maturation Plan 11 Task 3: Identification of Suitable Materials based on Literature Review, Computational Tools, and Lab Scale Measurements 13 Task 4: Production / Procurement of Suitable Materials, Construction of Lab Scale Test Unit 16 Task 5: Laboratory-Scale Breakthrough and Cyclic Testing of Selected Materials to Determine CO2 Capture Performance 19 Task 6: Process Simulation 25 Task 7: Initial Technical and Economic Feasibility Study 30 Task 8: Lab Testing and Simulation of Process Intensification 33 Task 9: Complete Preliminary Design and Costing of the Pilot Scale Unit 34 Task 10: Pilot Unit Fabrication, Shipping, and Installation 38 Task 11: Pilot-Scale Unit Testing 64 Task 12: Final Techno-Economic Analysis 82 Task 13: State Point Data Table 96 Task 14: Technology Gap Analysis 97 Task 15: Environmental Health and Safety Risk Assessment 99 Lessons Learnt During TCM Testing 100 Summary 100 Bibliography & Cited References 102 2 LIST OF TABLES Table 3.1 CO2 Capacities and CO2-N2 Separation Factors for Different Materials 15 Table 5.1 Amounts of CO2 Adsorbed and Desorbed During Various Cycles 23 Table 6.1 Process Simulation Summary for a 550 MW SCPC Plant with CO2 Capture 29 Table 7.1 Comparison of various CO2 compression and liquefaction schemes 31 Table 7.2 Capital, parasitic power, LCOE, and the CO2 capture cost 32 Table 9.1 Cost summary for the test skid 35 Table 10.1 Vaporizer Performance Prediction 42 Table 10.2 Key process equipment information 44 Table 10.3 TCM Acceptance Test Methodology 55 Table 11.1 Test matrix for parametric testing 64 Table 11.2 Variable values for parametric testing 65 Table 11.3 Operating conditions for the laboratory test unit 75 Table 11.4 Summary of performance of laboratory liquid regeneration unit 80 Table 12.1 Parasitic power requirements for the InnoSepra process 82 Table 12.2 Parasitic power requirements for the InnoSepra process based on TCM testing 83 Table 12.3 Summary of main streams – InnoSepra Process 84 Table 12.4 Updated energy requirements based on field testing 85 Table 12.5 Key results from process simulation 92 Table 12.6 Capture process integrated with a gas turbine 93 Table 12.7 TEA summary for Case B12A 93 Table 12.8 TEA summary for Case B12B-Derate 93 Table 12.9 TEA summary for Case B12C (InnoSepra Capture) 94 Table 12.10 TEA summary for Case B12D (InnoSepra Capture with GE LM7600 PG) 94 Table 12.11 Summary of various TEA Cases 95 Table 13.1 State Point Data Table 96 3 LIST OF FIGURES Figure 2.1 Block Flow Diagram of the Proposed CO2 Capture Process 12 Figure 3.1 Effect of Modification level on CO2 and N2 capacities for ZA1 14 Figure 3.2 CO2 adsorption isotherms for modified ZA1 sorbent 15 Figure 4.1 Lab test unit for the evaluation of various materials 17 Figure 4.2 Lab test unit for breakthrough testing 18 Figure 4.3 Apparatus for measuring CO2 solubility in various liquids 19 Figure 5.1 CO2 breakthrough curves for the ZA2 sorbent 21 Figure 5.2 Schematic of the laboratory unit 22 Figure 5.3 Concentration profile at the bed exit (adsorption step) 23 Figure 6.1 Flowsheet for process simulation 28 Figure 8.1 CO2 breakthrough curve for the structured sorbent 34 Figure 9.1 InnoSepra pilot plant schematic 36 Figure 9.2 3-D equipment layout & integration with the host plant 36 Figure 10.1 PFD for the feed preparation section 39 Figure 10.3 PFD for the liquid regeneration section 39 Figure 10.3 Details on the vaporizer heat exchanger 41 Figure 10.4 Schematic of the breakthrough test unit 43 Figure 10.5 Regenerant concentration at the exit of the adsorbent bed 43 Figure 10.6 3-D view of skids connected together 46 Figure 10.7 Plan view of various skids with dimensions 46 Figure 10.8 Final P&ID of skid 1 47 Figure 10.9 Final P&ID of skid 2 48 Figure 10.10 Final P&ID of skid 3 48 Figure 10.11 Three Photos of skids at the fabrication site 49 Figure 10.12 Three photos of skids prior to shipment 50 Figure 10.13 Cause and Effect Diagram 53 Figure 10.14 Snapshot of glycol regen system 54 Figure 10.15 Layout for the third site at TCM 56 Figure 10.16 Electric supply to the third site at TCM 57 4 Figure 10.17 Unloading of Skid 1 at TCM 58 Figure 10.18 Views of Skids 1 and 2 58 Figure 10.19 Views of Skids 1, 2, and 3 59 Figure 10.20 Schematic of the analytical skid 60 Figure 10.21 View of the installed bench scale unit 61 Figure 10.22 InnoSepra skids at TCM 62 Figure 10.23 InnoSepra skids at TCM 63 Figure 11.1 Performance of the InnoSepra Purification bed for the removal of NO2 from flue gas 66 Figure 11.2 Purification performance data (October 22, 2023) 66 Figure 11.3 InnoSepra dryer bed performance for moisture removal from flue gas 67 Figure 11.4 Dryer bed performance on November 10, 2023 68 Figure 11.5 300 nm3/h CO2 breakthrough curve, cooling water to the heat exchanger off 69 Figure 11.6 400 nm3/h CO2 breakthrough curve, cooling water to heat exchanger on 70 Figure 11.7 500 nm3/h CO2 breakthrough curve, cooling water to heat exchanger on 71 Figure 11.8 TCM regenerant heating test 72 Figure 11.9 Temperature evolution of CO2 adsorber with hot air flow at 60 nm3/hr 73 Figure 11.10 Schematic of the laboratory unit 75 Figure 11.11 Laboratory Unit 76 Figure 11.12 CO2 and temperature profiles during the test (internal heat exchanger) 77 Figure 11.13 CO2 concentration in the product/vent gas 77 Figure 11.14 Laboratory unit with external heat exchanger 78 Figure 11.15 Temperature and pressure profiles during the test (external heat exchanger) 79 Figure 11.16 CO2 profiles during the test (external heat exchanger) 79 Figure 11.17 CO2 concentration in the product/vent gas (external heat exchanger) 80 Figure 12.1 AspenPlus flowsheet of CO2 Capture test unit 84 Figure 12.2 InnoSepra CO2 capture process schematic and its integration with the 5 power plant 87 Figure 12.3 The InnoSepra Process integrated with the power plant 89 Figure 12.4 ASPEN Plus simulation for a 650 MWnet SCPC plant with CO2 capture (Case B12C) 91 Figure 12.5 ASPEN Plus simulation for a 650 MWnet SCPC plant with CO2 capture (Case B12E) 92 Figure 14.1 Flowsheet for Technology Gap Analysis 97 Figure 15.1 Key Waste Streams 99 6 This work was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or any third party’s use or the results of such use of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof or its contractors or subcontractors. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof, its contractors, or subcontractors. 7 SUMMARY InnoSepra’s project, “Transformational Sorbent-Based Process for a Substantial Reduction in the Cost of CO2 Capture,” utilized computational tools, materials characterization, and lab and pilot scale testing to optimize previously identified materials to determine their performance under post-combustion capture conditions. InnoSepra utilized the test results to determine the energy required for regeneration and to develop a process design/analysis to demonstrate the application of developed materials for post-combustion capture. A techno-economic analysis was performed to fully assess the potential of the materials for post-combustion capture. InnoSepra also updated the State Point Data Table and completed the environmental, health, and safety (EH&S) Risk Assessment. The InnoSepra CO2 capture technology has the potential for a significant reduction in the CO2 capture cost for the power plant and industrial flue gases. • It is possible to obtain very high recovery (90-95%) and high purity (>98%) CO2 with physical sorbents while meeting the EOR/sequestration product specifications. • The technology has the potential to reduce the thermal parasitic power by more than 65%, and the capital required by about 45% leading to about 45% reduction in the CO2 capture cost for the coal-based power plant flue gas. • After demonstration at the pilot scale and further process optimization, the process has the potential for a capture cost of around $40/MT with no increase in Levelized Cost of Electricity (LCOE) and with very little loss in power output. 1.1 Major Project Goals The project goals for InnoSepra’s project, “Transformational Sorbent-Based Process for a Substantial Reduction in the Cost of CO2 Capture,” were to demonstrate, through materials selection and lab-scale testing, that a transformational CO2 capture technology based on physical sorbents can help achieve DOE’s CO2 capture program goals in the short term. The project was to utilize computational tools, bench-scale testing, and detailed engineering design to estimate equipment sizes and parasitic power needs, and a techno-economic analysis to fully assess the potential of the process. Because the InnoSepra process uses materials with low heats of sorption, and heat extraction occurs at a much lower temperature than in solvent-based processes, the impact 8 on power loss due to steam extraction is significantly lower. Based on a preliminary techno- economic analysis, the InnoSepra team was to confirm projections that the process can lead to more than 50% reduction in capital cost (excluding CO2 compressor), and more than 70% reduction in power loss due to steam extraction relative to the leading solvent-based technologies such as Shell’s Cansolv solvent technology (Case B12B in DOE’s 2019 Baseline Report). The objectives of this project were to demonstrate that the InnoSepra process can achieve a projected CO2 capture cost below $40/tonne. A pathway to achieve a capture cost below $30/tonne was to be provided as well. After lab-scale testing of the process, another objective of this project was to integrate all the process components and test with an actual flue gas source in support of the engineering study and the techno-economic analysis. Yet another objective of this project was to provide a Technology Maturation Plan to show a pathway toward achieving these objectives. The ultimate goals of the Project were to confirm the expected performance at the lab and bench-scale and to provide sufficient data to enable a detailed techno-economic analysis for a commercial-scale plant, confirming the projected economics. This was to provide a high degree of confidence in the applicability, cost-effectiveness, and practical feasibility of this process and demonstrate that this technology can meet the overall performance goal for the development of advanced CO2 capture and separation technologies (90% CO2 removal at a capture cost of less than $40/tonne and preferably below $30/tonne) and limit the increase in LCOE by more than 30% compared to the current approaches. 1.2 Progress Against Goals Work was completed on all the tasks for the accomplishment of project goals and objectives. During this project work on tasks related to the Project Management Plan (Task 1), Technology Maturation Plan (Task 2), Identification of Suitable Materials based on Literature Review, Computational Tools, and Lab Scale Measurements (Task 3), Production/Procurement of Suitable Materials, and Construction of Lab Scale Test Unit (Task 4), Laboratory-Scale Breakthrough and Cyclic Testing of Selected Materials to Determine CO2 Capture Performance (Task 5), Process Simulation (Task 6), Initial Technical and Economic Feasibility Study (Task 7), Lab Testing and Simulation of Process Intensification (Task 8), Bench-Scale Unit Process Design and Costing (Task 9), Bench Unit Fabrication, Shipping and Installation (Task 10), Bench-Scale Unit Testing 9 (Task 11), Final Techno-Economic Analysis (Task 12), State Point Data Table (Task 13), Technology Gap Analysis (Task 14) and Environmental Health and Safety Risk (Task 15) was carried out. The details are provided in the sections below. Based on a combination of lab and field testing, process simulation, and techno-economic analysis, the CAPEX for the InnoSepra Process for an SCPC Plant is about 35% lower than Shell Cansolv. The CO2 capture cost for the InnoSepra Process is about 45% lower than Shell Cansolv. CO2 capture using Shell Cansolv leads to a 23% reduction in the power output, while the reduction in the power output for the InnoSepra Process is about 14%. The reduction in the power output for the InnoSepra Process can be as low as 2.5% if an NGCC plant is used to provide the electrical power for capture. These results are in line with the project goals in the previous section. The project was carried out in two budget periods. Tasks 1-9 were carried out during Budget Period 1, and the Key Milestones for Budget Period 1 (about 24 months) were: 1. Identify suitable materials (sorbent and regenerant) for the proposed process 2. Carry out lab-scale testing and process simulation 3. Preliminary technical and economic feasibility study 4. Design & costing of the pilot-scale unit Tasks 10-15 were carried out during Budget Period 2, and the Key Milestones for Budget Period 2 (about 36 months) were: 1. Fabrication and installation of the pilot-scale test unit 2. Commissioning and testing of the pilot-scale unit 3. Detailed engineering of the full-scale capture plant 4. Final Technical and Economic Feasibility Study 5. Technology gap analysis 6. Completion and presentation of the closeout report 10 1. ACCOMPLISHMENTS Task 1.0 – Project Management and Planning This task included all work elements required to maintain and revise the Project Management Plan and to manage and report on activities in accordance with the plan. It also included the necessary activities to ensure coordination and planning of the project with DOE/NETL and other project participants. These included, but were not limited to, the submission and approval of required NEPA documentation. The Project Management Plan, including the updated GANTT chart, updates to the Statement of Project Objectives (SOPO), milestones, and deliverables, was finalized during the first budget period. A poster and an oral presentation on the project were made at the DOE CO2 Capture Meeting in Pittsburgh in August 2019. A project update to DOE was provided during the Kickoff Meeting in September 2019. Due to cost and contract issues with Arizona State University, it was decided to measure the sorption isotherms in-house. It was also decided to use Adroitech (started by an Inventys cofounder) for Monte Carlo simulations and structured sorbent fabrication. Due to travel and workplace restrictions related to COVID-19, Adsorptech of New Jersey was chosen for the bench unit design, P&IDs, costing, and HAZOP. A meeting with DOE and Technology Centre Mongstad (TCM) was held in August 2020 to discuss bench-scale testing at TCM. Several pieces of information related to materials and vendor evaluation were provided to TCM for these activities. These included Material Safety Datasheets and reports on the Technology Maturation Plan, utilities required for testing, and tie-ins with the host plant. InnoSepra’s internal HAZOP was provided to TCM, and a HAZOP for the interface with the test site was carried out with TCM. Both the Test Agreement and HAZOP were completed in Q1-21. Task 2 – Technology Maturation Plan The objectives of this task were to prepare and submit a TMP in accordance with the SOPO, showing the Technology Readiness Levels (TRLs) of the components of the InnoSepra technology, along with details of how the project will advance the TRLs and what future work will 11 be required to continue TRL advancement. Results and Discussion: A detailed Technology Maturation Plan was provided to DOE. A block flow diagram of the InnoSepra Process is shown in Figure 2.1. The flue gas is purified in unit #1, moisture is adsorbed in unit #2, CO2 is adsorbed in unit #3, CO2/regenerant mixture is separated in unit #4, regenerant is heated in unit #5, and residual regenerant is adsorbed in unit #6. The starting TRL of unit #1 was 6, and at the end of the project, the TRL was 7. The starting TRL of unit #2 was 5-6, and at the end of the project, the TRL was 7. The starting TRL of unit #3 was 4- 5, and at the end of the project, the TRL was 4-5. The starting TRL of unit #4 was 8-9 (commercial technology) and did not change during the project. The starting TRL of unit #5 was 8-9 (commercial technology) and did not change during the project. The starting TRL of unit #6 was 8-9 (commercial technology) and did not change during the project. Regenerant CO2 LP STM In Depleted FG #5 HE Cond Out #6 Adsorption #3 CO2 Adsorption CO2 to Compression Regenerant/CO2 Mixture #4 Condenser #2 Moisture Adsorption #1 FG Purification From DCC Figure 2.1: Block Flow Diagram of Proposed CO2 Capture Process 12 Task 3 – Identification of Suitable Materials based on Literature Review, Computational Tools, and Lab Scale Measurements The objectives of this task were to identify suitable sorbent materials (4-5) for lab-scale testing based on literature review, predictions based on computational tools, and measurement of identified material properties using the adsorption microbalance. A detailed literature search was carried out to identify sorbent materials and regenerants suitable for the proposed process. The materials identified through the literature search were further narrowed based on Monte Carlo simulations and measurements of sorption isotherms in the adsorption microbalance. The identified regenerants needed to have low CO2 solubility at both the regeneration temperature as well as the adsorption temperature. Finally, based on environmental considerations, the chosen regenerants needed to have a low global warming potential and low ozone depletion potential. Results and Discussion: The results of the literature search indicated that commercially available sorbents, as well as their modified forms, will be suitable for the process. This will provide materials with different pore sizes. The absolute CO2 capacities for the identified materials range between 14 and 17-wt% (at 25oC and a CO2 partial pressure of 0.15 atma), and the CO2/N2 selectivities range between 15- 20 for the commercial materials and over 100 for the modified materials. The two most likely regenerants included liquids with boiling points between about 95 °C and 120 °C. Monte Carlo simulations and sorption isotherm measurements confirmed the literature search results. The effect of modifications on CO2 and N2 capacities for one commercial material, as predicted by Monte Carlo simulation, is shown in Figure 3.1. As the modification level approaches about 20% the material has essentially no N2 capacity, and the loss in CO2 capacity is about 25%. At higher modification levels, the material has a very high CO2/N2 selectivity of 660 and a moderately lower CO2 capacity compared to commercially available material. 13 Figure 3.1: Effect of modification levels on CO2 and N2 capacities for a commercial material A procedure for modifying commercial materials was developed. The CO2 & N2 isotherms as a function of temperature (0, 20, 40, 60, 80, 100oC) were measured on these materials using Micromeritics ASAP-2020. Sorption isotherms for CO2 and N2 on commercial materials and modified materials are in line with literature values as well as Monte Carlo simulations. One of these isotherms is shown in Figure 3.2. For a commercial material with 50%-modification, a CO2 capacity of about 17-wt% for a CO2 partial pressure of about 150 mm Hg (15% CO2 in flue gas at 1-atm) can be obtained. The summary of CO2 capacities and CO2-N2 separation factors is provided in Table 3.1. 14 Table 3.1: CO2 Capacities and CO2-N2 Separation Factors for Different Materials 25 20 15 Q, g/100g 10 5 0 0 100 200 300 400 500 600 700 800 P, mm Hg Figure 3.2: CO2 adsorption isotherms at 300C for the modified sorbent Various regenerants were tested with the selected sorbents to ensure that a cyclic steady state was obtained without any significant regenerant or sorbent degradation. Based on this criterion, two regenerants, EM1 and EM2, were identified. These regenerants had no ozone depletion potential; however, they did have a significant global warming potential. Additional regenerants were identified to address TCM’s concerns regarding the potential emissions of the regenerants 15 used in the process regeneration. Two of these regenerants, EM3 and EM4, have no ozone depletion potential and a very low global warming potential. Based on additional discussions with TCM, including the HS&E (Health, Safety & Environmental) evaluation, EM4 was chosen for pilot testing. Task 4: Production / Procurement of Suitable Materials, Construction of Lab Scale Test Unit The objective of this task was to produce/procure all the materials identified in Task 3 and build a lab-scale test unit for their evaluation. The lab-scale test unit was to have the capability to generate simulated flue gas (primarily a CO2/N2 mixture, 5-10 slpm flow) with the ability to add other flue gas components, such as moisture, SO2, and NO2. The test unit was to be fully automated and to have the ability to control adsorption (25-40 oC) and regeneration temperatures (90-110 oC). Commercially available materials were obtained and modified as discussed in Task 3. The lab-scale test unit for the evaluation of various materials is shown in Figure 4.1. It consists of two beds for CO2 adsorption and regeneration/regenerant-CO2 separation. The lab-scale test unit has the capability to generate simulated flue gas (primarily a CO2/N2 mixture, 5-10 slpm flow) with the ability to add other flue gas components, such as moisture, SO2, and NO2. The test unit is fully automated and has the ability to control adsorption (25-40 oC) and regeneration temperatures (90-110 oC). In addition to the lab test unit shown in Figure 4.1, a two-bed breakthrough test unit shown in Figure 4.2 was fabricated. This unit can carry out adsorption tests at temperatures up to 40 oC to determine breakthrough capacities and can be regenerated with nitrogen at temperatures up to 120 o C. 16 Figure 4.1: Lab test unit for the evaluation of various materials 17 Figure 4.2: Lab test unit for breakthrough testing The apparatus for solubility measurements is shown in Figure 4.3. It consists of a stirrer with a heating plate, a means to deliver gas to the system, and instruments to measure the pressure and the temperature in the system. After the liquid has come to the desired temperature, a known quantity of the gas is introduced, and both the pressure and the temperature are monitored to determine the steady state. The solubility is determined from the known amounts of gas and the liquid in the system. 18 Figure 4.3: Apparatus for measuring CO2 solubility in various regenerants The CO2 solubility measurements were carried out for the chosen regenerant (EM4) at various CO2 partial pressures and temperatures. For a CO2 partial pressure of 0.2 atm at 25oC, the CO2 solubility was low, similar to the CO2 solubility in water. Task 5: Laboratory-Scale Breakthrough and Cyclic Testing of Selected Materials to Determine CO2 Capture Performance The objectives of this task were to conduct lab-scale testing with up to 4 selected materials under a variety of conditions in breakthrough as well as cyclic modes to determine CO2 capacities, CO2 purities, and recoveries, and the energy needed for regeneration. Breakthrough tests with two materials (ZA1 and ZA2) at 25oC using a feed containing 15% CO2 in N2 were carried out to enable sorbent downselection. 19 Results and Discussion: The results for the ZA2 sorbent are shown in Figure 5.1 and indicate that the breakthrough times and the CO2 capacities at the expected adsorption temperature (~25oC) are significantly higher than the breakthrough times and the CO2 capacities at the expected regeneration temperature (~100 oC). The results for the ZA1 material were similar to those for the ZA2 material, with much higher capacities at the expected adsorption temperature (~25oC) and lower capacities at the expected regeneration temperature (~100 oC). In general, CO2 capacities for the ZA1 material are lower than those for the ZA2 material, and the breakthrough curve is more spread out compared to the breakthrough curve for the ZA2 material. All the sorbents tested during Budget Period 1 have very good CO2-N2 separation factors, 26- 654, as indicated in Table 3.1. These separation factors are more than adequate to produce high- purity CO2 (>95%) in the InnoSepra Process. Based on breakthrough testing and isotherm measurements carried out during this budget period, the ZA2 material has the highest net CO2 capacity (difference in the capacity between adsorption and regeneration) and a very good CO2- N2 selectivity. The regeneration energy consists of the following: Regeneration Energy = Heat of Adsorption + Energy Needed to Heat the Sorbent + Energy Needed to Heat the Vessel + Sensible heat of CO2 20 Figure 5.1: CO2 breakthrough curves for the ZA2 material The heat of adsorption for all the sorbents is nearly the same, about 38 kJ/mol of CO2. Because of the highest net CO2 capacity, the ZA2 material will have the lowest amount of total energy needed, as it needs the lowest amount of energy to heat the sorbent and the vessel. For this reason, the ZA2 material was chosen for further testing. The regenerants with the lowest ozone depletion potential and global warming potential, and low CO2 solubility, EM3 and EM4, are suitable for use with the ZA2 material. The regeneration tests were carried out with the regenerant in both liquid and vapor form. Later tests were carried out with regenerant vapor, as this approach can reduce the regenerant flow by a factor of more than 20 compared to liquid regenerant. This is due to the fact that the latent heat of evaporation is significantly higher than the liquid sensible heat. EM4 was used for these tests at a temperature of 110 oC. We were able to reduce the liquid flow by a factor of over 5, and the vapor phase regeneration gave a higher CO2 capacity compared to regeneration with liquid alone. For regeneration tests with EM4 vapor, a synthetic mixture of CO2 and N2 was prepared by 21 mixing pure streams of the individual gases, and the concentration of feed CO2 was varied between 6 and 12%. The mixture was fed to the adsorbent bed at a flow rate between 7.5 and 10 slpm until the CO2 concentration at the bed exit was 1-2 % At that point, the regenerant was fed counter- currently to the feed at a flow rate of 0.5 to 1 slpm for about 7 min. This was followed by another 7-minute step in which the regenerant flow was stopped, and a vacuum pump was turned on to complete the removal of the adsorbed carbon dioxide and any regenerant remaining in the bed. The bed was then cooled down with a nitrogen stream to be ready to start the cycle again and receive the dry flue gas stream. Several experiments were carried out to ensure that results were reproducible and that the system reached a steady state, namely, the amounts adsorbed and desorbed converged, and there was no CO2 accumulation in the beds. Cyclic testing Figure 5.2. Schematic of the laboratory unit A schematic of the laboratory unit is shown in Figure 5.2. The unit consists of a bed containing ZA2 material, which has a high carbon dioxide adsorption capacity. The bed is 5/8” in diameter 22 and 60” long and contains 1/16” pellets of the adsorbent material. The regenerant is EM4, liquid at room temperature. It is stored in an adjacent vessel, which is kept hot by recirculating it through an inline heater. A typical graph of the CO2 concentration during the adsorption step is shown in Figure 5.3. CO2 (%) 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 2 4 6 8 Figure 5.3. Concentration profile at the bed exit (adsorption step), x-axis (time in mins), y-axis (% CO2 concentration) Several experiments were carried out to ensure that results are reproducible and that the system reaches a steady state, namely, the amounts adsorbed and desorbed converged, and there was no CO2 accumulation in the beds. A summary of the amounts of CO2 adsorbed and desorbed for some of the cyclic tests is provided in Table 5.1. It can be seen that these amounts do not change significantly with time. In some cases, the amounts desorbed are greater than those adsorbed. This can be due to the fact that the amounts adsorbed are determined from the breakthrough curve (adsorption step) while the amounts desorbed are measured with a totalizer. Table 5.1. Amounts of CO2 Adsorbed and Desorbed During Various Cycles Test 1 - Feed Flowrate: 10 lpm - CO2 concentration: 6% Regeneration Flow Rate: 1 lpm – Temperature: 105 oC CO2 adsorbed CO2 desorbed Cycle # (liters) (liters) 1 5.3 6.3 2 4.2 5.4 3 5.0 5.8 4 5.0 5.8 23 Test 2 - Feed flowrate: 10 lpm - CO2 concentration: 6% Regeneration flowrate: 1 lpm – Temperature: 105 oC CO2 adsorbed CO2 desorbed Cycle # (liters) (liters) 1 7.4 2 5.1 5.6 3 6.1 5.6 4 6.3 5.5 5 6.2 5.6 6 6.0 5.6 7 6.0 5.6 8 6.0 5.7 Test 3 - Feed flowrate: 10 lpm - CO2 concentration: 6% Regeneration flowrate: 1 lpm – Temperature: 105 oC CO2 adsorbed CO2 desorbed Cycle # (liters) (liters) 1 7.35 2 4.16 5.84 3 5.39 5.58 4 5.42 5.58 5 5.56 5.6 6 5.54 5.59 7 5.51 5.6 8 5.44 5.83 9 5.44 5.6 10 5.50 5.53 11 5.42 5.63 12 5.41 5.46 13 5.60 5.53 14 5.55 5.63 15 5.51 5.69 16 5.51 5.79 24 Test 4 - Feed flowrate: 7.25 lpm - CO2 concentration: 12% Regeneration flowrate: 1 lpm – Temperature: 105 oC CO2 adsorbed CO2 desorbed Cycle # (liters) (liters) 1 5.45 5.40 2 5.38 5.70 3 5.36 5.40 4 5.35 5.40 5 5.35 5.80 6 5.34 5.50 7 5.37 5.90 8 5.34 5.50 Test 5 - Feed flowrate: 6.0 lpm - CO2 concentration: 12% Regeneration flowrate: 1 lpm – Temperature: 105 oC CO2 adsorbed CO2 desorbed Cycle # (liters) (liters) 1 4.93 5.80 2 4.94 5.60 3 4.98 5.50 4 5.80 5.60 5 4.90 5.50 6 4.89 5.30 7 4.93 5.25 8 4.94 5.65 For all these tests, the average product purity was >98.5%, and the CO2 recovery was greater than 92%. While it is very difficult to get an accurate estimate of the energy required for regeneration in a small lab-scale test (much higher heat losses than a bench scale or a pilot scale unit, higher vessel mass to CO2 product flow ratio), the estimated energy required for regeneration (based on energy input to the heater) was below 2.0 GJ/MT. We expect this number to be lower in the pilot scale tests. Task 6: Process Simulation The objective of this task was to update the process model based on the results of lab-scale testing. The deliverable from this task was a process model that combined that could be used to predict the performance of full-size CO2 capture plants, for process optimization, and for sizing 25 various process equipment. A process simulation model integrating the CO2 capture process and the coal-based power plant was developed to take advantage of potential synergies. Aveva’s ProII process simulation software was licensed for this work. Results and Discussion: The initial work was focused on optimizing the CO2 compression and liquefaction parts of the capture. The starting point is compressed CO2 after the 4th compression stage. The CO2 stream at 45 bara is obtained after a 4-stage compression of CO2 from the CO2 capture process. This gaseous CO2 stream is cooled in the compressed CO2 cooler before being sent to the CO2 condenser. The liquid CO2 is pumped to a pressure of 150 bara in the process. The ProII model for the process, including four compression stages and the stream table for the process, was provided in the quarterly reports. A techno-economic analysis is provided under Task 7. A complete process simulation was set up in Aveva Pro/II to evaluate the performance of the CO2 recovery and compression process for a 550 MW power plant. The feed (74,092 kmol/hr) enters the process at 57 oC and 100 kPa, and its composition is 68.1% N2, 13.5% CO2, 15.17% H2O, 8.2% Ar, and 0.02% O2. This flow rate is very close to the flow rate in Case B12A of the 2015 DOE Baseline Report (Cost and Performance Baseline for Fossil Energy Plants, Vol 1a: Bituminous Coal (PC) and Natural Gas to Electricity, Rev 3, DOE/NETL-2015/1723, July 6, 2015). The feed in Figure 6.1 first goes to a SO2 removal unit (EX-SO2_REM), its pressure is raised to 114.1 kPa, and it is then cooled down to 27.8 oC in a direct contact cooler unit (T1). The resulting stream enters a PSA dryer (EX_DRYER) in which water vapor is adsorbed, and the dry stream enters the CO2 adsorption beds (EX_CO2_BEDS) containing the CO2 sorbent. The adsorbed CO2 is desorbed in two stages; in the first stage, a hot regenerant flows down the bed to heat it and facilitate the CO2 desorption, and, in the second stage, a vacuum is applied to remove an additional amount of CO2 as well as any regenerant remaining in the bed. The CO2/regenerant stream is sent to a tank (T-102), where gas/liquid separation takes place, and the resulting stream is reheated to 100oC to be used during the regeneration step for the CO2 beds. The rich CO2 stream is cooled 26 down to 10 oC to condense vapor from the regenerant and then enters an adsorbent bed where any remaining regenerant vapor is adsorbed. The product stream now contains >99.0% CO2 and less than 0.02% O2. This stream is compressed to 5,580 kPa through a four-stage compression (C1 to C4). The compressed gas goes through two heat exchangers, E1 and E2, to condense it and is pumped to 15,270 kPa for pipeline transport. The product flow rate is 9,516.5 kmol/hr, which represents a 95% recovery of CO2. A summary of the utilities required for the process flowsheet in Figure 6.1 is provided in Table 6.1. 27 Figure 6.1. Flowsheet for Process Simulation 28 Table 6.1: Process Simulation Summary for a 550 MW SCPC Plant with CO2 Capture 29 Task 7 – Initial Technical and Economic Feasibility Study The objective of this task was to prepare and submit an initial TEA in accordance with the SOPO. The initial TEA utilized the information from the lab-scale testing and process simulations (Tasks 5-6). Results & Discussion: Two separate TEAs were carried out. In the first TEA, the benefits of the proposed CO2 compression scheme were compared with the conventional CO2 compression cycle. In the second TEA, the overall capital and energy required for CO2 capture were estimated to determine the CO2 capture cost. For the first TEA, it was assumed that the product was 100% CO2 to enable comparison with various cases. In reality, the product will have up to 1% N2 + Ar. This assumption, however, is not expected to have any material impact on the results of this study. Other assumptions are listed below: 1. Based on the DOE feedback, a power cost of $60/MWh and a scaling factor of 0.61 for the compressor cost were used. 2. To simplify the calculation of the CO2 capture costs, a capital charge of 19.5% to convert capital cost to an annualized cost was used (2% for the annual levelized maintenance cost and 17.5% capital charge factor) for the TEA. The updated Technoeconomic analysis for the CO2 compression process is provided in Table 7.1. Case 1, the InnoSepra Base Case, corresponds to Case 1 in a DOE report (Development of CO2 Purification and Compression System Spreadsheet Tool, DOE/NETL-4001/080415, August 2015). Case 2, InnoSepra Process, is based on the process of Figure 6.1 with 4-stage gaseous compression followed by liquid pumping. Case 3, Amine Base Case, corresponds to Case 4 of the DOE report and is representative of a solvent-based capture process such as the Linde-BASF process (CO2 product at 3.4 bara, 102oC). The final case, Case 4, corresponds to an amine base case with 3-stage compression and glycol drying to produce dry CO2 at 45 bara. This is followed by CO2 condensation and liquid pumping as in Case 2. The CO2 product flow (on a dry basis) is 11,500 kmol/hr in all cases. 30 Table 7.1: Comparison of Various CO2 Compression and Liquefaction Schemes 31 Potential energy savings of about 19% for the InnoSepra process and about 20% for the Linde- BASF type process are possible compared to Case 1. The cooling duty for Case 2 (InnoSepra with Liquid Pumping) is about one-third of the Amine Base Case (Case 3). Table 7.2: Capital, Parasitic Power, LCOE, and the CO2 Capture cost Case 1 Case 2 Case 3 Case 4 MEA InnoSepra InnoSepra InnoSepra Gen 2 Capital Capture Gen 1 Gen 2 + NGCC Feed Preparation, ID Fan, DCC, $MM $35 $28 $28 CO2 capture system, $MM $124 $105 $91 CO2 compression + intercoolers, $MM $37 $40 $37 Letdown turbine, engg, shipping, installation $50 $42 $42 Total, $MM $397 $246 $215 $198 Utilities LP steam@110 C, lbs/hr 700,000 568,750 568,750 Power equivalent of steam, MW 87 32 24 24 Cooling water, kgal/hr 24,000 11,500 10,000 10,000 Electric power, MW 67 67 55 35 Total parasitic power, MW 154 99 79 59 % of plant output 28 18 14.4 11 Cost of Utilities With and Without Capture Steam without capture, $/1,000 lb 5.83 5.83 5.83 5.83 Electric power without capture, $/MWh $64 $64 $64 N/A Electric power for NGCC, $/MWh $55 LCOE and the CO2 Capture Cost Increase in LCOE, % 75 48 38 0 Amount of CO2 captured, MM tonnes/yr 3.18 3.18 3.18 3.18 Capital charge per year, $MM $77.4 $48.0 $41.9 $38.6 Annual cost of parasitic power, $MM $82.9 $53.3 $42.5 $27.3 Labor + sorbent replacement/disposal, $MM $20.0 $6.2 $6.2 $6.2 Cooling Water, $MM/yr $17.5 $8.4 $7.3 $7.3 Total Annual Cost, $MM $198 $116 $98 $79 CO2 capture cost at the plant gate, $/tonne $62 $36 $31 $25 TS&M, $MM/yr $14 $14 $14 $14 Total Annual Cost including TS&M, $MM $212 $130 $112 $93 CO2 capture cost including TS&M, $/tonne $67 $41 $35 $29 CO2 capture system = HX + dehyd vessels + dryer skid + sorbent + CO2 vessels + skid + sieve + vacuum 32 The TEA for the overall process is provided in Table 7.2. This TEA assumes a sorbent life of 5 years based on prior experience. The results in Table 7.2 indicate that a capture cost of $35/tonne (including TS&M) is possible for the proposed InnoSepra process (Case 3). If the electrical energy for capture (power consumption for pumps, blower, vacuum pump & compressor, and the power lost due to steam extraction from the SCPC plant) is provided by an NGCC plant at $55/MW, then a capture cost below $30/tonne is possible with no increase in LCOE. In this simplified TEA for Budget Period 1, the capital cost is converted to an annualized cost using a capital charge factor. The TEA for the second budget period was more detailed and was based on Appendix D of the FOA. Task 8 – Lab Testing and Simulation of Process Intensification The objective of this task was to test process intensification at the lab scale using a structured version of the CO2 capture sorbent identified in Task 5 for testing under various process conditions. The key objective of this task was to determine the increase in sorbent productivity (sorbent volume required/flue gas processed) compared to particulate sorbents with a pressure drop similar to or lower than the particulate sorbents. CO2 Breakthrough Curve 7 6 5 4 3 2 1 0 0 5 10 15 20 25 Time, minutes Figure 8.1: CO2 breakthrough curve for the structured sorbent (6% CO2 in N2, 10 slpm, 25oC), % CO2 on the y-axis 33 Structured sorbents in the form of sheets containing ZA2 material were fabricated by Adroitech during the first Budget Period after resolving the initial coating issues. A test unit to evaluate these sorbents was built. The unit had the capability to control inlet and outlet flows, compositions, and temperatures. The sorbent was activated following the procedure suggested by Adroitech. However, the breakthrough tests showed premature breakthroughs and a significantly lower CO2 capacity compared to tests with the sorbent in pellet form. A typical breakthrough curve is shown in Figure 8.1. A likely explanation is shrinkage during activation, leading to flow bypass along the wall. The sorbent structure was sent back to Adroitech during Q1-21 for diagnosis and repair. Based on our discussions with Adroitech, the sorbent structure was repaired, and the breakthrough data were provided. The breakthrough front was sharper, but the capacity was too low to be useful. Task 9.0 – Complete Preliminary Design and Costing of the Bench Scale Unit The objective of this task was a preliminary design and costing of the process unit capable of processing 40-100 scfm of flue gas. Long lead time items were also identified so that they could be ordered early in Budget Period 2. Another objective of this task was to come up with the costs for the bench scale test unit. Most of the work during this budget period was focused on P&IDs for the bench-scale unit, process and HS&E documentation for TCM, and evaluation of the integration options for the InnoSepra test unit at TCM’s new plant test site. The test unit consists of three skids: - Feed preparation skid - Adsorption skid - Regeneration skid The overall PFD for the test unit is shown in Figure 9.1. All the switching valves are 6” butterfly valves. The cost summary for the test unit is provided below. 34 Table 9.1: Cost Summary for the Test Skid Item Description Key Components Overall Cost Feed Preparation Skid Feed blower (20 HP motor w/ $65,000 VFD), purification bed Adsorption Skid Dryer and CO2 adsorption beds, $85,000 vacuum pump (5 HP) Regeneration Skid 100 kW heater, 10 HP liquid $90,000 pump, vapor-liquid separator with condenser Electrical System & PLC $50,000 Shipping $15,000 Skid Fabrication $290,000 Sieve, heat transfer liquid Sorbents for drying, CO2 $40,000 adsorption Non-InnoSepra Engineering $60,000 Total $695,000 The total budgeted cost for test unit design and fabrication during the two budget periods was about $800,000, excluding shipping (Table 9.1). A significant amount of design and HAZOP work was done during Budget Period 1. The 3-D layout for the integration of the InnoSepra test unit with the TCM site is shown in Figure 9.2. The MSDS for all the materials used in the InnoSepra process were provided to TCM for their evaluation of health and safety risks associated with the InnoSepra Process. An internal InnoSepra HAZOP for the test unit was completed, and a copy was provided to both TCM and DOE. The InnoSepra HAZOP was reviewed with TCM. Extensive information for TCM’s HS&E evaluation was provided. Most of the HS&E work consisted of providing responses to TCM regarding chemicals used in the InnoSepra Process and emissions from the InnoSepra Process. 35 To Flue Gas Duct Heat Exchanger D-202A D-202B Adsorber To Vaccum Adsorber C.W. in Pump HE-306 C.W. out VP-303 Vacuum Pump HE-304 Regenerant Regeneration Heater (Electric) VFD Motor D-201A D-201B Dryer CO2-Depleted Dryer Process Flue Gas to T-305 Water Duct VP Suction KO Drum KV100 C-102 Packing Feed HE-103 Blower DCC Flue Gas Feed T-101 Condensate Separator D-100 Purification Bed VFD Motor BPR To Seawater To Seawater Return Line Return Line P-105 Figure 9.1: InnoSepra Pilot Plant Schematic Figure 9.2: 3-D Equipment Layout and Integration with the Host Plant 36 Key Milestones for Budget Period 2 (BP2) 7. Fabrication and installation of the pilot scale test unit 8. Commissioning and testing of the pilot-scale unit 9. Detailed engineering of the full-scale capture plant 10. Final Technical and Economic Feasibility Study 11. Technology gap analysis 12. Complete and present the closeout report Key Activities for BP2 Skid design and testing, field test report, and updated TEA. Most of the effort focused on pilot unit fabrication, including lab tests in support of skid design: • Capable of processing up to 500 nm3/hr of flue gas • Designed as three separate skids • Each skid is about 8’ w x 10’ H and 25’ L, about 12,000 lbs each • Very significant challenges due to fabrication resources, engineering resources, and supply chain constraints, leading to project delays Pilot Unit Testing • All the major components were procured and sent to the fabricators • All the skids were completed, and initial testing was done at the fabrication site • All the skids were shipped to TCM • The skids were installed, commissioned, and tested 37 BP2 Tasks Task 10 – Pilot Unit Fabrication, Shipping, and Installation The objective of this task was to complete a detailed engineering design of the test unit and get the test unit fabricated, shipped, and installed at TCM. Another goal of the test unit design was to minimize emissions during testing at TCM A detailed engineering design of the test unit was carried out with the goal of minimizing the footprint during shipping and installation, and for ease of operation during testing. The overall PFD of the test unit is shown in Figure 9.1. The test unit was designed as three separate skids. Skid 1 consisted of the following sections: feed purification for the removal of condensate as well as NO2, SOX, and aerosols, feed compression and cooling, and feed drying. Skid 2 consisted of four adsorption beds for CO2 separation from the flue gas. The process cycle consists of: adsorption, regeneration, cooling, and repressurization. The four-bed cycle allows continuous operation. Skid 3 consists of the following sections: vacuum pump for evacuation, heating of regeneration medium, and cooling of CO2 product. Some highlights of the PFD are: - A condensate separator, T-101, is placed prior to the D-100 purification bed to prevent any freezing in the purification bed - A packed column Direct Contact Cooler (DCC) with structured packing is used to cool the compressed flue gas - Process water is used in the DCC as the amount of water required is fairly small. The process water from T-101 and DCC will be returned to the seawater return line at TCM. - A knockout drum before the vacuum pump VP-303 is used to reduce the load on the vacuum pump - A two-stage cooling is used to separate the regenerant and carbon dioxide. The initial cooling is done with process water to about 20 oC (HE-306), and the final cooling is done with a freon chiller (HE-307). 38 - A guard bed D-307 for the final capture of the regenerant vapor The approach outlined ensures that minimal regenerant emissions will occur during the testing. The PFDs for the feed preparation and regeneration sections are shown in Figures 10.1 and 10.2. Figure 10.1: PFD for the feed preparation section Figure 10.2: PFD for the regeneration section 39 Regenerant Heater / Vaporizer One of the concerns that arose during the detailed design in Q1-2022 was the sizing of the regenerant heater/vaporizer HE-304. We rigorously modeled this heat exchanger using the data provided by one of the vendors (CEXI). The calculation indicated that the heat exchanger did not have enough surface area to vaporize the required quantity of EM4. After we shared our results with the vendor, the vendor withdrew the bid. A quote was also received from Graham Manufacturing (Helicoil heat exchanger). The vendor uses a proprietary design, and there was no way to verify their design using the available tools. It was decided to develop our own design tools for this vaporizer using rigorous methods (Chen, Correlation for Boiling Heat Transfer to Saturated Fluids in Convective Flow, I&EC Pro Des Dev, v5, n3, p322, July 1966) as well as the Aveva ProII simulator. The results from both methods were within 25% of each other. Using these tools, we were able to model a commercially available heat exchanger from McMaster-Carr and an existing InnoSepra heat exchanger (Figure 10.3). The simulation results indicate that using two McMaster heat exchangers in parallel or using one McMaster heat exchanger in series with the existing InnoSepra heat exchanger (Figure 10.3) will provide enough heat transfer area to vaporize 100% of EM4. Further lab testing coupled with simulation indicated that the existing InnoSepra heat exchanger is sufficient to vaporize 100% of EM4, provided EM4 flows on the tube side.. Only about 60% of the tube area is needed for EM4 vaporization. The remaining area is used for superheating. 40 Figure 10.3. Details on InnoSepra’s vaporizer heat exchanger Upon the expiration of the license of the Aveva Pro/II process simulation software, it was decided to switch simulation software provider. A contract signed with AspenTech gave InnoSepra the license to AspenPlus, HySys, and Aspen Adsorption. Several cases that had been run with Pro/II were set up in AspenPlus or HySys to verify that the same or similar results were being generated. For the current project, the first step was to verify the vaporizer performance (heat exchanger HE-304). Since the regenerant used in the process is not in the software databases, it was necessary to either create a new compound or modify a similar compound. Due to issues with modifying the temperature-dependent parameters in AspenPlus, the same thermodynamic model was not used for both software. The modified Soave-Redlich-Kwong (SRK) model had been used in Pro/II, while the Non-Random Two-Liquid (NRTL) model was used in AspenPlus. The results obtained are summarized in Table 10.1. 41 Table 10.1. Vaporizer performance prediction AspenPlus REG IN REG OUT PG IN PG OUT Mass Flow kg/hr 2480 2480 10014 10014 Temperature oC 80 115.4 140 128.4 Molar Vapor Fraction 0 0.984 0 0 Overall heat tr.coeff. W/m2-K 131 Aveva Pro/II REG IN REG OUT PG IN PG OUT Mass Flow kg/hr 2465 2465 10000 10000 Temperature oC 80 126.0 140 127.5 Molar Vapor Fraction 0 1.000 0 0 Overall heat tr.coeff. W/m2-K 123 It can be seen that, while Aveva Pro/II predicts complete vaporization, AspenPlus predicts only about 98% vaporization. The exit temperature of the heating media is very similar in both cases, and the overall heat transfer coefficient is slightly higher in AspenPlus. The differences observed can be attributed to the different thermodynamic models and heat transfer correlations used. Guard bed design The pilot plant, designed to be installed at TCM, includes an adsorbent bed downstream of the regenerant recovery system to remove any vapor from the product stream. In order to properly size the adsorber, a laboratory unit was built to carry out breakthrough tests. The carrier gas flows into a bubbler filled with the regenerant at a low temperature so that its concentration is about the same as expected during pilot testing (Figure 10.4). This mixture flows through the adsorbent bed, and the vapor concentration is measured both before and after the adsorber using a Bacharach Gas Monitor. The breakthrough tests were carried out using N2 or CO2 as the carrier gas. No significant difference in the adsorption capacities was observed using either gas. A typical breakthrough concentration profile for a bed containing a commercial sorbent is shown in Figure 10.5. The adsorber was designed to keep the exit concentration below 10 ppm throughout the whole testing period at TCM. 42 Figure 10.4. Schematic of the breakthrough test unit Figure 10.5. Regenerant concentration at the exit of the adsorbent bed Three different sorbents were tested, and no significant difference in capacities was observed. A commercially available sorbent was used in the guard bed. 43 Engineering Design and Fabrication The design package for RFPs, including the CO2 capture section and the regenerant handling section, was sent to three fabricators (Plant Process Engineering of Houston, Gamlet Inc. of York, PA, and Xytel, Inc. of Roebuck, SC) in Q4 2021. A requirement for factory acceptance testing was also included in the RFPs. Only one bid was received. The bid was significantly over budget, and after including missing items needed to fabricate the test unit, the fabrication cost would have exceeded the budget by at least $500k. Based on this response, the design and project management were brought in-house. InnoSepra, with help from Adsorptech, performed detailed process and mechanical design, skid design, and control system design of the entire process, including the costing of major process components. Procurement from multiple sources (vessels, rotating machinery, control valves, instruments, heaters, heat exchangers, sorbents, analyzers, PLC, I/O cards, liquid heater, and freon chiller) was done to expedite fabrication. The detailed process design of all the major process components included: - Specifications for all the process components - Detailed mechanical design of all the vessels completed - Identified components on the critical path The key process equipment suppliers and reception date are shown in Table 10.2. Table 10.2: Key Process Equipment Information Component Description Supplier Fabrication/Reception date T-101 Condensate Separator Gamlet Q3 2022 D-100 Purification Vessel InnoSepra unit Q3 2022 D-201A and D-201B Dryer Gamlet Q2 2022 Vessels D-202A and D-202D CO2 Vessels Gamlet Q2 2022 T-305 Suction Drum Gamlet Q3 2022 T-301 Liquid Tank Gamlet Q3 2022 D-307 Carbon Bed Gamlet Q3 2022 C-102 Feed Blower Hibon Q1 2022 VP-303 Vacuum Pump Edwards Q1 2022 P-103 Liquid Pump Grainger Q1 2022 44 P-105 Liquid Pump Grainger Q1 2022 P-302 Liquid Pump Sunair Q1 2022 R-306 Freon Chiller ACME Q2 2022 Switching valves J.T. Lynn Q2 2022 PLC, I/O Cards Automation Direct Q2 2022 HE-304 Heat Exchanger InnoSepra unit 2022 HE-304 Liquid Heater Durex Industries Q3 2022 HE-103 DCC Koch-Glitsch Q1 2022 HE-306 Heat exchanger Xchanger Inc. Q1 2022 HE-307 Heat exchanger Xchanger Inc. Q1 2022 Analyzers Horiba, ThermoFisher, 2022 Bacharach, Manalytical Process instruments Miller Energy Q2 2022 Insulation of the vessels and pipes, as well as the installation of the electrical cables, was completed in the 1st quarter of 2023. Also, the heat exchangers were installed, and the different adsorbents were loaded onto the respective vessels. Pressure testing of the different vessels in the plant and connecting pipes was done using pressurized nitrogen. The purification bed D-100 was loaded with the previously identified material. A commercially available sorbent was used in the dessiccant beds. The adsorber beds D-202 were loaded with the ZA2 material, and a sorbent for regenerant vapor was loaded in the guard bed. A 3D drawing of the skids is shown in Figure 10.6, and a plan view with dimensions is shown in Figure 10.7. 45 Figure 10.6: 3-D view of various skids connected together Figure 10.7: Plan view of various skids with dimensions 46 The final P&ID was released during Q2 2023, and the pages corresponding to the 3 skids are shown in Figures 10.8 to 10.10. The skids prepared for shipping are shown in Figures 10.11 and 10.12. Figure 10.8. Final P&ID of Skid 1 47 Figure 10.9. Final P&ID of Skid 2 Figure 10.10. Final P&ID of Skid 3 48 Figure 10.11: Three Photos of skids at the fabrication site 49 Figure 10.12: Three photos of skids prior to shipment 50 51 HAZOP Analysis Regular meetings were held with TCM to ensure compliance with site codes related to fabrication and operation (electrical and mechanical). An HAZOP analysis was carried out for the whole plant. A summary of changes from the initial analysis is as follows: Node 1 (Duct to feed blower) - The prepurification bed is choked, and the Condensate buildup in T-101 is choking the flow Action: Periodic check on the condensate level, and manual opening of the valve. - Residue Fluid Catalytic Cracking (RFCC) Flue gas micron-size particulate in the gas can choke the D-100 bed. Action: Monitor DPI (Pressure Drop) Node 8 (Low-temperature Glycol circuit) - The valve to the Dryer bed exchanger fails to open or any other block discharge. Action: Install a new pressure transmitter, not in the P&IDs - The Variable Frequency Drive (VFD) on the glycol pump is reversed. Action: Recommend pressure monitoring on P401 discharge - Block discharge of the pump Action: Install a new pressure transmitter, not in the P&IDs - The valve to the Dryer bed exchanger fails to open or any other block discharge. Action: Install a new pressure transmitter, not in the P&IDs Node 9 (High-temperature Glycol circuit & Vaporizer) - The VFD on the glycol pump is reversed. Action: Recommend pressure monitoring on P401 discharge - Heat exchange tubes in the HE304 vaporizer break or develop leakage. Action: Recommend a pressure transmitter to be installed in the Glycol pump discharge to see a large drop in pressure due to tube failure. Update operating instructions to include inspection daily. - Water in the system instead of glycol. 52 Action: Update operating instructions to include a weekly inspection to check Glycol quality. - FICV 403 fails to open and causes block discharge. Action: Install a new pressure transmitter, not in the P&IDs Node 10 (Water Cooler, Chiller, and vapor capture) - Heat exchanger tubes in HE-301A break or develop a leak Action: Recommend checking the quality of regeneration fluid periodically in the operating procedure. It is recommended to dye the regen liquid to be able to determine leakage. Based on the HAZOP analysis, a Cause & Effect diagram was generated (Figure 10.13). It summarizes the alarms and trips that will follow any deviation from the normal operation of the plant. Figure 10.13. Cause and Effect diagram 53 As the list of trips and alarms was finalized, procedures for different types of shutdowns were developed. In a Normal Shutdown, CO2 beds are shut down, and Bed A is manually pressurized with Instrument Air (IA). Also, Dryer beds, Hot Glycol/Regenerant/Low Temp Demineralized Water (DMW), Feed, and Utilities (Process water, IA, Chiller) are shut down. In a Warm Shutdown, only the Dryer beds and CO2 beds are shut down. In a Soft Shutdown, CO2 beds continue cycling to complete the step they are in. Then, CO2 beds, Dryer beds, Hot Glycol/Regenerant/Low Temp DMW, and Feed are shut down. The Utilities are kept running for 1 or 2 hours, and then everything is shut down. In an emergency shutdown, everything is stopped. The software to operate the plant was developed by Adsorptech, and a snapshot of one of the screens is shown in Figure 10.14. The different units of Skid 3 can be seen: Separator, Vacuum pump, regenerant, Glycol Heater, Vaporizer, Heat Exchangers, Chiller, and Guard Bed. An emergency stop button, as well as different alarms that could be triggered during the operation of the process, are also shown on the screen. Figure 10.14. Snapshot of High-Temperature Glycol Regen screen Acceptance test A test methodology was developed for the bench scale unit (Table 10.3). Each skid will be 54 tested independently as soon as they are completed. There will be a final master test when all three skids are connected. This approach was established to expedite the testing and acceptance process by TCM. Table 10.3. TCM acceptance test methodology TCM ACCEPTANCE TEST METHODOLOGY ELECTRICAL MEGGER CHECKS PROCEDURE: PRIOR TO POWER-UP OF ELECTRICAL PANEL, MEGGER ALL MOTORS & BUBARS (PHASE TO GND). DISCONNECT MOTORS FROM VFD AND STARTERS AND REMOVE FUSES FROM FUSE HOLDERS. MEGGER USING FLUKE 1507 INSULATION TESTER AT 1000V FOR 1 MINUTE. ELECTRICAL POWER-UP & CHECK PROCEDURE: AFTER MEGGER TESTS, CONNECT TO GENSET (480V, 480AMP, 3 PHASE SEREVIE) & POWER-UP OF INDIVIDUAL CIRCUITS & CHECK. ELECTRICAL SETUP PROCEDURE: SET MOTOR OVERLOAD TO REQUIRED FLA. SETUP VFD PARAMETERS PER VENDOR INSTRUCTIONS, REFLECTING MOTOR OPERATING PARAMETERS AND IO CONFIGURATION. MOTOR DIRECTION CHECKS PROCEDURE: CHECK MOTOR ROTATION OF ALL MOTORS. FUNCTION CHECK ESTOP PROCEDURE: POWER-UP ESTOP CHECK ESTOP CIRCUIT. VERIFY FAULT SIGNAL FUNCTION. FUNCTION CHECK IO (DI) PROCEDURE: POWER UP PLC & HMI, AFTER LOADING PROGRAM VERIFY EACH DIGITIAL INPUT. CONFIRM PLC IO AND INDICATION ON HMI. FUNCTION CHECK IO (D0) PROCEDURE: POWER UP PLC & HMI, AFTER LOADING PROGRAM VERIFY EACH DIGITIAL INPUT. CONFIRM PLC IO AND INDICATION ON HMI. PRESSURIZE INSTRUMENT AIR SYSTEM & SETUP FILTER REGULATORS AT REQURED PRESSURE. FUNCTION CHECK IO (NETWORK) PROCEDURE: POWER UP PLC & HMI. RUN EACH VFD VIA THE HMI AND CONFIRM NETWORK IO FUNCTION FUNCTION CHECK IO (AI) PROCEDURE: FUNCTION CHECK ANALOG INPUT AS PER DETAILS 55 Shipping and Installation TCM proposed that the InnoSepra unit be installed at the 3rd site at Mongstad. The layout is shown in Figure 10.15, and the available electrical supply is shown in Figure 10.16. It was concluded that both space and electrical supply were sufficient for the InnoSepra skids. Figure 10.15: Layout for the third site at TCM 56 Figure 10.16: Electric supply to the third test site at TCM Quotes to ship the three skids and a 40 ft. container with spare parts to TCM, Mongstad, were requested from Bring Cargo (Norway) and ProFreight (USA). After analysis of the quotes received, Bring was chosen as the freight forwarder. A Carnet was applied for and issued to facilitate the temporary export/import of the plant components. The 3 skids and container were picked up from the fabrication site in York, Pa, in the first week of June 2023, and were loaded on two different ships. The cargo arrived at TCM Mongstad in early July. The process of unloading the skids and positioning them in TCM’s 3rd site is shown in Figures 10.17 to 10.19. 57 Figure 10.17. Unloading of Skid 1 at TCM Figure 10.18. View of Skids 1 and 2 58 Figure 10.19. View of Skids 1, 2, and 3 Three crates with analytical equipment and mechanical spare parts were air freighted in early July 2023, and the installation and calibration of the CO2 and NO2 analyzers were performed in August. A schematic of the analytical skid is shown in Figure 10.20. 59 Figure 10.20. Schematic of the analytical skid The installation of the skids and pre-commissioning activities were carried out during July 2023. InnoSepra, with help from Adsorptech, carried out the commissioning of the units during August. The installed test unit is shown in Figure 10.21. 60 Figure 10.21. View of the installed test unit The pretesting of the units started in September. Several components in the skids, such as the VFD controlling the vacuum pump and the contactor associated with the vaporizer, did not work properly and had to be replaced. The installed test skids are shown in Figures 10.22 and 10.23. Further tuning of various control loops was carried out during the quarter. The major issues encountered during commissioning included the ability to heat and vaporize the regenerant and recover it in two heat exchangers (HE301A – cooling water heat exchanger; and HE301B – low-temperature chiller). A significant amount of time was spent on tuning the freon chiller that provides low-temperature cooling for HE301B. The design temperature of the freon chiller is -30 oC. However, we were not able to get a temperature below -10 oC during the tuning.. TCM had the requirement to validate the regenerant emissions analyzer against TCM’s FTIR analyzer before any gas could be sent out of the system. Once TCM was able to validate the vendor calibration for the regenerant analyzer with a TCM-generated standard, we were able to run all the skids together. 61 Extended operation of the flue gas purification units and dryers was possible as these unit operations are not dependent on CO2 bed heating with the regenerant. Breakthrough testing of CO2 beds with dry flue gas was also carried out. The beds were regenerated with instrument air after the breakthrough tests. Figure 10.22: InnoSepra Skids at TCM 62 Figure 10.23: InnoSepra Skids at TCM 63 Task 11 – Pilot-Scale Unit Testing The objectives of this task were to test the complete system at TCM, analyze the test data, and decommission the test unit. The goal of these tests was to demonstrate that the technology can meet CO2 recovery (>90%), CO2 purity (>95%), and parasitic power (<1.8 GJ/MT) targets during extended testing. Testing of the test unit at TCM was carried out in two stages. In the first stage, while the emissions analyzer validation was taking place, the purification and dryer beds were tested by running flue gas through those units at a set flowrate (300 nm3/hr), and the purified/dry flue gas was sent back to the return flue gas duct. In the second stage, after completion of the emissions analyzer validation, testing of the CO2 removal process was carried out. The test matrix for parametric testing is provided in Table 11.1. The corresponding values of the variables are shown in Table 11.2. Table 11.1. Test matrix for parametric testing Variables Test 1 Test 2 Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Test 9 Test 10 Test 11 Test 12 Test 13 Test 14 CO2 concentration in flue gas H H H H H H H H H H H H H H Flue gas flowrate H H M M H H H H H H H H H H Dryer beds half cycle time M M M M M M M M M M H H M M Heating /Cooling ratio CO2 adsorbers cycle time M M M M H H M M M M M M L L Vacuum level M M M M M M L L M M M M M M Fluorocarbon flowrate M M M M M M M M H H M M M M Variables Test 15 Test 16 Test 17 Test 18 Test 19 Test 20 Test 21 Test 22 Test 23 Test 24 Test 25 Test 26 Test 27 Test 28 CO2 concentration in flue gas H H H H H H H H H H H H H H Flue gas flowrate H H M M M M M M M M M M H H Dryer beds half cycle time M M M M M M M M H H M M M M Heating /Cooling ratio CO2 adsorbers cycle time M M H H M M M M M M L L M M Vacuum level M M M M L L M M M M M M M M Fluorocarbon flowrate L L M M M M H H M M M M M M L: Low value M: Medium value H: High Value 64 Table 11.2. Variable values for parametric testing Variables Low Medium High CO2 concentration in flue gas (%) 4 7 10 Flue gas flowrate (Nm3/h) 300 400 500 Dryer beds half cycle time (min) 8 10 12 Heating /Cooling (ratio) 0.15/0.85 0.2/0.8 0.3/0.7 CO2 adsorbers cycle time (min) 6 8 10 Vacuum level (mbar - abs) 200 250 300 (in Hg vacuum) 23.6 22.1 20.7 Regenerant flowrate (kg/h) 2000 2500 3000 Flue Gas Purification A collection of the NO2 removal performance data from the Mongstad Heat Plant (MHP) and RFCC flue gases utilizing InnoSepra’s flue gas purification technology was carried out. The MHP flue gas has 8-9% CO2. NO2 is an impurity in the flue gas and would end up in the CO2 product for the InnoSepra Process. NO2 is also a catalyst for solvent degradation in solvent-based capture processes. Successful performance would be very beneficial since it can lead to a significant reduction in solvent degradation and loss. For most tests, the flue gas flow rate was set to 300 nm3/hr (FT103), while a few additional tests were conducted at 400 nm3/hr and 500 nm3/hr. Bed temperature and pressure were the same as the flue gas supply temperature and pressure. The NO2 analyzer was calibrated and validated against TCM’s online NO2 measurement. NO2 concentration in the feed varied between 12 and 20+ ppm, and NO2 in the flue gas exiting the purification bed was continuously measured. The results confirm that the purification bed was performing satisfactorily, and NO2 was removed to below the analyzer detection limit, <0.03 ppm, for the entire test period. Figure 11.1 shows the performance for the period spanning from October 2023 to January 2024 for MHP flue gas. All of the feed NO2 was removed during the entire test 65 period, while little or no NO (Nitric Oxide) was removed. Only NO2 is known to cause amine degradation, and NO removal is not required to prevent this. Not having to remove NO provides a significantly lower NOX removal cost for the InnoSepra Process compared to other technologies, such as SCR (selective catalytic reduction), that remove both NO and NO2. Figure 11.1. Performance of InnoSepra Purification bed for removal of NO2 from MHP flue gas The bench scale unit purification performance data during a 24-hour period is shown in Figure 11.2. The flue gas flow rate was 300 nm3/hr. Figure 11.2. Purification performance data (October 22, 2023) 66 More than 800 hours of testing was performed for the removal of NO2 utilizing InnoSepra’s flue gas purification technology. Dryer Bed Performance Testing of InnoSepra’s sorbent-based dryer technology, which is an integral part of InnoSepra’s CO2 capture technology, was carried out. Moisture removal is needed to produce a dry CO2 product from the CO2 beds since incoming flue gas was moisture-saturated at 10-15 oC. Data for the dryer bed performance was collected for over 250 hours, and the results indicate satisfactory performance with moisture removal to below 5 ppm levels. The moisture removal performance for the period spanning from October 2023 to January 2024 is shown in Figure 11.3. Figure 11.3. InnoSepra dryer bed performance for moisture removal from MHP flue gas Figure 11.4 shows the dryer bed performance on November 10, 2023. The flue gas flow rate was 300 nm3/hr, and the moisture in the flue gas was reduced from 1000+ppm to <10 ppm in less than 4 hours, and a steady state was reached. 67 Figure 11.4 – Dryer bed performance on November 10, 2023 (MHP Flue Gas) Except for the bed pressure drop, no external source of energy is needed for the sorbent regeneration. In addition to the InnoSepra Process, this drying technology is applicable to other capture technologies. CO2 Breakthrough Curves The original test plan was to run the 4 adsorber beds on a cycle, but due to the issues associated with the regenerant heating and a leak in one of the heat exchangers, the plan was modified to run single-bed adsorber tests to collect relevant CO2 capture data. As part of the effort, the primary focus was to identify any issues and improve the heating performance of the adsorber bed with the regenerant. Tests were performed to get the CO2 breakthrough curves, which provide essential information for the operation of the test unit. Tests were done at three different flue gas flow rates: 300, 400, and 500 nm3/hr. All tests utilized MHP flue gas. The 300 nm3/hr test was performed with 9.5% CO2 in the feed. The 400 nm3/hr and 500 nm3/hr tests were performed with 8.5% CO2 in the feed. The CO2 breakthrough curves are shown in Figures 11.5 to 11.7. For the 300 nm3/hr test, there was no cooling water flowing through the heat exchanger. No heat was removed during adsorption, and there was no fluctuation in the outlet CO2 concentration. On the other hand, cooling water did flow through the heat exchanger for the 400 nm3/hr and 500 nm3/hr tests. The removal of some of the heat of adsorption leads to fluctuation in the bed outlet CO2 concentration. 68 For the 300 nm3/hr test, the concentration and temperature fronts closely follow each other (Figure 11.5). If the adsorption stopped at about 30-32 minutes, virtually all the heat of adsorption would stay in the bed, and the heat of adsorption would not be needed during regeneration. Retaining the heat of adsorption does increase the average adsorption temperature and reduces the CO2 capacity. However, the savings from retaining the heat of adsorption, 0.6-0.8 GJ/MT, significantly outweigh the additional heat needed to heat the additional sieve and the vessel, <0.1 GJ/MT. For low-feed CO2 concentrations (<5% CO2), more than 75% of the heat of adsorption can be retained without a significant impact on capacity. For higher feed CO2 concentrations (>10% CO2), at least some heat of adsorption, 25-50%, would need to be removed to get acceptable CO2 capacities. Figure 11.5 – 300 nm3/h CO2 breakthrough curve, 9.5% CO2 in the feed 69 Figure 11.6. 400 nm3/hr CO2 breakthrough test 70 Figure 11.7. 500 nm3/hr CO2 breakthrough test As expected, as the feed flow rate increases, the outlet temperature at the end of the test is higher. At the same time, the outlet concentration increases faster at the breakthrough point, and as pointed out before, it then fluctuates for the higher feed flowrates (400 and 500 nm3/hr). The analysis of the experimental data collected during the adsorption step indicates that the CO2 capacity measured at different breakthrough points is in line with expectations and matches well with laboratory testing under the same conditions. Heating requirements Tests were carried out to determine the time required to heat the CO2 beds with liquid regenerant. The results show that the bed can be heated quickly with the regenerant (Figure 11.8), and the operation is reproducible, having been repeated more than five times. 71 Figure 11.8. TCM regenerant heating test Figure 11.9 shows the evolution of the adsorber bed inlet and outlet temperature as hot instrument air flows through the bed from top to bottom. Only after about 2.5 hours does the bed bottom see the heat front advancing. This information will allow us to evaluate the energy requirements to heat the system when there is no adsorption. 72 Figure 11.9. Temperature evolution of the CO2 adsorber with hot air flow at 60 nm3/hr Operational Issues Regenerant analyzer calibration: Not finalized until December 2023; only skids 1 and 3 could be tested until mid-October, and only closed-loop heating for skids 2 and 3 until mid-December. Chiller Operation: A low-temperature chiller (HE301B) was installed to condense most of the regenerant vapor at -30 oC before it goes to the guard bed. Despite many attempts to fix the chiller, we were not able to operate the chiller below -5 oC. Regenerant Heating: Regenerant heating issues in HE-304 made it very difficult to transfer enough heat to the CO2 beds for regeneration. We were able to get temperatures over 100 oC in both Skid 2 (CO2 beds) and Skid 3 (regenerant heating) in manual operation mode. However, this could not be repeated. An alternative regenerant was identified and tested after the HS&E evaluation. It was loaded in the Hot Liquid Tank, but only limited tests were carried out. Hot air regeneration after the adsorption step was used to determine the energy needed for regeneration. In a single-bed cycle, the adsorption data (capacities, recoveries) can be obtained independently of the regeneration cycle. Freezing issues: Due to the unusual and continuous low ambient temperatures sustained during Winter at Mongstad, several freezing episodes were observed that affected the operation of the 73 plant. The plant was not designed to operate in the freezing weather, and several sections of the plant had to be thawed, and the purification bed had to be emptied to prevent further issues. Emissions & Environment: There were no adverse emissions incidents during the testing period based on emissions monitoring. A small amount of coolant did spill due to overflow. This was, however, within the permissible limits. The operation of all the unit operations together could not be demonstrated at TCM due to the reasons listed above. Various unit operations in the process were demonstrated either individually or in combination with other components. This includes: • Flue gas purification: >800 hrs • Flue gas drying: >250 hrs • CO2 breakthrough testing, about 15 experiments • Thermal energy needed for regeneration (air heating): 8 experiments • Heating with an alternate regenerant to demonstrate liquid heating Decommissioning The test unit was shut down on March 15, 2024. Decommissioning of the test unit was carried out by InnoSepra and TCM personnel. The various skids were disconnected, and all the connecting parts, cable trays, etc., were loaded into the 40 ft. shipping container. InnoSepra personnel packed and air-shipped the analyzers and the control laptop. All three skids, along with the analytical skid, were wrapped for shipment back to the U.S. Laboratory test unit After the TCM testing was completed, laboratory tests were carried out to obtain purity and recovery data that could not be obtained at TCM. The lab unit closely reproduces the CO2 adsorption process implemented in the pilot plant unit installed at TCM. The CO2 adsorber is a 1.5” diameter 4-ft long bed containing 1164.5 g ZA2 sorbent. A schematic and a photo of the unit are shown in Figures 11.10 and 11.11. The feed, a mixture of CO2/N2 with a CO2 concentration of 10.2%, flows upward through the bed during the adsorption step, and the top product concentration is monitored with a Horiba 510-VIA CO2 analyzer. EM4 is used as the regenerant. It is heated up to a temperature of 105-110 oC and, during the regeneration step, flows downward through the bed and then through a cooling coil and condenser to ensure all 74 the liquid vapor is condensed back into the hot liquid tank. The gas out of the condenser flows through a guard bed to remove trace amounts of remaining regenerant and then to the CO2 analyzer. The cycle steps, step times, and flow rates used during this period are summarized in Table 11.3. These tests have the objective of optimizing both the recovery and purity of the CO2 product. Table 11.3. Operating conditions of the laboratory test unit Step Duration Gas/Liquid Flowrate Flow direction (min) (l/min) Adsorption 10-15 Feed 20-25 upward Regeneration 10-15 Liquid Regenerant 0.5-2 downward Drain 0.5-2 N/A N/A downward Vacuum 10-15 N/A N/A downward Cooling 5-15 Nitrogen 5-30 upward/downward Figure 11.10. Schematic of the laboratory unit 75 Figure 11.11. Laboratory test unit The CO2 and temperature profiles for the tests carried out show that a cyclic steady state is achieved after 5-6 cycles (Figure 11.12). The CO2 concentration profile obtained with this bed configuration is shown in Figure 11.13. During the adsorption step, the measured concentration is from the bed top, and the breakthrough occurs at about 12.5 minutes. For the remaining steps, the CO2 concentration corresponds to the bottom product stream. During the vacuum step, the concentration is above 90%, with a maximum value of 98.5%. The recovery of CO2 is greater than 90%, most of it occurring during the vacuum step. 76 Figure 11.12. CO2 and temperature profiles during the test Time, hr:mm:sec Figure 11.13. Cycle steps for one of the cycles in Fig. 11.16 (CO2 concentration in the product/vent gas) An external heat exchanger was considered as an alternative for cooling the gas during both the adsorption and cooling steps. The heat exchanger is connected to the packed bed at a height of about 2/3 of the total bed height. This has some advantages because it can lead to a simpler design of the CO2 adsorption beds in the full-scale unit. A photo of the unit is shown in Figure 11.14. 77 As with the previous configuration, the results show that the CO2 concentration is above 90% during the vacuum step, with a maximum value of 98.3%. The recovery of CO2 is greater than 90%, most of it occurring during the vacuum step. Although reasonably good performance is obtained, the bed is not cooled down to the extent required, and the bed capacity is affected. Figure 11.14. Laboratory unit with external heat exchanger The temperature and pressure profiles during these tests are shown in Figure 11.15, and the concentration profiles are shown in Figure 11.16. The CO2 concentration profile obtained with this bed configuration is shown in Figure 11.17. During the liquid regeneration step, a much lower CO2 concentration is seen coming out at the bottom of the bed. 78 Figure 11.15. Temperature and pressure profiles during the test (external heat exchanger) Figure 11.16. CO2 profiles during the test (external heat exchanger) 79 Time, hr:mm:sec Figure 11.17. CO2 concentrations for a full cycle in the product/vent gas A summary of the laboratory results from the tests carried out between April and June 2024 is shown in Table 11.4. Tests done between April 3 and May 21 correspond to the original bed configuration, with the internal heat exchanger. Tests between May 30 and June 10 are with the external heat exchanger configuration. Table 11.4. Summary of performance of the laboratory liquid regeneration unit Date 3-Apr 4-Apr 12-Apr 17-Apr 18-Apr 19-Apr 1-May 2-May 7-May 13-May 15-May 17-May 20-May 21-May 30-May 3-Jun 4-Jun 10-Jun Adsorption flow, lpm 21 21 25 25 25 25 21 21 21 21 21 21 21 21 18 18 18 18 Feed CO2, % 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 10.2 Cooling flow, lpm 29.7 29.7 29.7 22.5 22.5 22.5 22.5 19 9.5 9.5 9.5 5 5 5 5 5 5 5 Step # Cycle times: 1 adsorption time, min 12 12 10 10 10 10 12 12 12 12 12 12 13 11.5 10 10 10 10 2 liquid heating time, min 10 10 9 9 9 9 10 10 10 10 11 11 12.5 11.5 9.5 9.5 9.5 9.5 3 liquid drain time, min 2 2 1 1 1 1 2 2 2 2 1 1 0.5 0.5 0.5 0.5 0.5 0.5 4 vacuum regeneration time, min 12 12 10 15 12 10 12 12 12 12 12 12 13 12 10 10 10 10 6 cooling time, min 12 12 10 10 13 10 12 12 12 12 12 12 6 12 10 10 10 10 Vacuum step Average CO2 % 83.9 84.3 80.9 86.2 85.2 84.4 87.3 92.2 90.5 92.0 90.3 95.3 95.6 95.6 85.6 83.6 89.5 80.3 Max CO2 % 93.0 93.8 95.4 97.3 95.8 97.0 96.0 96.5 94.4 95.1 94.0 98.4 98.5 97.2 93.1 92.4 95.2 98.3 CO2 recovery in vacuum step, % 52.0 55.2 57.0 65.5 62.4 58.7 65.8 69.2 68.5 86.6 85.0 76.5 78.4 75.7 85.0 76.6 83.8 75.2 80 Laboratory and Field Testing Summary • The laboratory testing indicates about 90% recovery for a CO2 purity greater than 95%. This was determined by measuring flows and CO2 concentrations for all the streams in the process. Further process optimization will allow for improving both recovery and purity. Higher recovery is possible through the recycling of a small portion of the gas during the cooling step. • Further process optimization is underway to improve both the recovery and the purity • Through a combination of TCM testing, lab testing, and process modeling, we are able to provide a good estimate of the purity, recovery, and parasitic power for the SCPC power plant conditions • A complete process model has been set up in ASPEN Plus. The model has been updated using the lab and the TCM data • Based on TCM and lab testing, the changes needed for the reliable operation of the pilot plant have been identified. These may be tested in the future as part of the Technology Maturation Plan 81 Task 12 – Final Techno-Economic Analysis The objectives of this task were to prepare and submit a final TEA in accordance with FOA Appendix D. The techno-economic analysis was to be performed using DOE’s Rev 4 TEA guidelines (Cost and Performance Baseline for Fossil Energy Plants, Vol 1: Bituminous Coal and Natural Gas to Electricity, September 24, 2019, NETL-PUB-22638). Modeling of the InnoSepra process The modeling of the whole plant in Aveva’s ProII was carried out during BP1 to estimate the electrical and thermal energy required for the process. Measured breakthrough tests provide some insight into potential parasitic power savings by retaining the heat of adsorption. The initial heating estimates for the InnoSepra Process for this project are provided in Table 12.1. This assumes that all the heat of adsorption, 0.8 GJ/MT, is lost during the adsorption step. Testing at TCM, as well as previously testing, indicated that depending on how the CO2 bed is operated, a significant amount of the heat of adsorption can be retained in the bed and does not have to be provided during regeneration. This can be done by stopping the adsorption step before the heat front has exited the bed. This reduces the heat required for desorption to 0.2-0.4 GJ/MT, and the total thermal energy needed is between 1.0-1.4 GJ/MT at about 110oC (Table 12.2). Assuming an average heating requirement of 1.2 GJ/MT, this is about a 22% reduction over the base Case in Table 12.1. The effective parasitic power based on lost work analysis will be lower than this number since the steam for the InnoSepra Process is extracted at a much lower temperature compared to amines (steam extraction temperature of about 160 oC). Table 12.1: Parasitic Power Requirements for the InnoSepra Process Heat of Adsorption 0.80 GJ/MT Vessel + Sieve Heating 0.60 GJ/MT Miscellaneous 0.15 GJ/MT Dehydration 0.00 GJ/MT Total (excluding compression) 1.55 GJ/MT 82 Table 12.2: Parasitic Power Requirements for the InnoSepra Process Based on TCM Testing Heat of Adsorption 0.20-0.40 GJ/MT Vessel + Sieve Heating 0.60 GJ/MT Miscellaneous 0.20 GJ/MT Dehydration 0.00 GJ/MT Total (excluding compression) 1.0-1.4 GJ/MT At the end of the license period for the Aveva ProII software, a contract was signed with AspenTech that has given InnoSepra a license to AspenPlus, Hysys, and Aspen Adsorption. One of the drawbacks was that the regenerant used in the InnoSepra process is not in this software database. Following discussions with AspenTech, it was decided to use a similar component for the simulations, but modify the temperature-dependent and independent properties with the information given by the regenerant manufacturer. The modeling of the whole plant in AspenPlus was carried out, and the corresponding flowsheet is shown in Figure 12.1. The results of the simulation are summarized in Table 12.3. S1 represents the feed stream that goes to separator T-101 (to remove liquid water in the feed) and purification unit D-100. The resulting stream S2 goes to a blower that raises its pressure to 1.29 bara, and stream 3 enters the Direct Contact Cooler (DCC). There it comes in contact with process water stream S40, which cools it down to 21.7 oC. Stream 4 enters the dryers D-201, and the dry product stream S8 is fed to the adsorber beds D-202. The CO2 is desorbed using a hot regenerant, and the resulting mixed stream enters a hot liquid tank T-301. The gaseous stream is cooled down first with process water S60 in the Heat Exchanger HE-301A, and then with a heat exchanger/chiller system HE-301B. The regenerant is condensed and returned to the liquid tank while the gaseous stream S23 goes to a Guard Bed, where traces of the liquid regenerant are adsorbed. Stream S7 is the final CO2 product. 83 Figure 12.1. AspenPlus flowsheet of the CO2 Capture test unit Table 12.3. Summary of main streams – InnoSepra process Stream Name Units S1 S3 S4 S8 S7 S40 S60 From Feed C-102 DCC D-201 GUARDBED Process Water Process Water To B1 DCC D-201 D-202 Product DCC HX-303 Phase Mixed Phase Vapor Phase Vapor Phase Vapor Phase Vapor Phase Liquid Phase Liquid Phase Temperature C 20.0 46.3 21.7 21.7 20.0 15.0 15.0 Pressure bar 1.05 1.29 1.20 1.19 1.11 1.29 1.29 Molar Vapor Fraction 0.972 1 1 1 1 0 0 Molar Liquid Fraction 0.028 0 0 0 0 1 1 Mass Enthalpy cal/gm -438.6 -382.5 -383.4 -347.8 -2136.4 -3798.9 -3798.9 Mass Density gm/cc 0.0012 0.0014 0.0015 0.0015 0.0020 1.00 1.00 Enthalpy Flow cal/sec -86667 -74265 -74181 -66456 -62674 -339342 -1060230 Average MW 29.5 29.8 29.8 30.0 44.0 18.0 18.0 Mass Flows kg/hr 711.3 699.0 696.6 687.9 105.6 321.6 1004.7 Mole Fractions NITROGEN 0.801 0.824 0.826 0.844 0 0 0 OXYGEN 0.034 0.035 0.035 0.036 1.00E-05 0 0 CARBON DIOXIDE 0.106 0.109 0.108 0.110 0.99 0 0 WATER 0.050 0.022 0.022 0.001 0 1 1 ARGON 0.009 0.009 0.009 0.009 0 0 0 NOx/SOx 1.00E-04 0 0 0 0 0 0 84 The energy requirements were further updated based on field test results using the ASPEN Plus Process Simulator (Table 12.4). Table 12.4: Updated Energy Requirements Based on Field Testing Heat of Desorption * 0.20-0.40 GJ/MT Vessel + Sieve Heating 0.75 GJ/MT Miscellaneous 0.20 GJ/MT Dehydration 0.00 GJ/MT Total (excluding compression) # 1.15-1.35 GJ/MT Notes: * A significant amount of heat of adsorption stays in the bed and does not need to be provided during regeneration. # This is the absolute amount of energy needed, not corrected for the lower steam extraction temperature. Integration of the InnoSepra Process with a Power Plant The 2019 Baseline Report (James et al., Cost and Performance Baseline for Fossil Energy Plants, Volume 1: Bituminous Coal and Natural Gas to Electricity, NETL-PUB-22638, September 24, 2019) was used for the techno-economic analysis. The two cases from the report relevant to capture from an SCPC plant are: Case B12A – Base case with 650 MWnet, no CO2 capture Case B12B – Coal flow to the plant is increased to increase steam and power generation for CO2 capture with Shell Cansolv (DOE Baseline Solvent). The plant output, 650 MWnet, is maintained. The InnoSepra cases include: Case B12B-Derate – Case B12A with CO2 capture using Shell Cansolv. The steam and power are extracted from the Base Plant, and there is no plant upsizing. 85 Case B12C – Capture using the InnoSepra Process. The steam and the power are extracted from the Base Plant. Case B12D – Case B12C with makeup power provided by a GE LM6000 PG gas turbine. No CO2 is captured from the gas turbine. Case B12E – Case B12D with CO2 capture from both the coal plant and the gas turbine The InnoSepra Process that forms the basis of this project and its integration with a pulverized coal-fired power plant are shown in Figure 12.2. In Figure 12.2, the additional unit operations needed for CO2 capture are shown in blue, new cooling water lines are shown in green, and new steam lines are shown in red. The pressure of the flue gas before the FGD is raised using a fan/blower by about 2-2.5 psig. The flue gas exiting the FGD, typically containing 50-100 ppm SOX, is sent to a direct contact cooler (DCC) to cool it to between 25 0C and 35 0C. Some of the SOX in the flue gas is removed in the direct contact cooler through the use of sodium hydroxide. The flue gas stream exiting the DCC is sent to a flue gas purification stage wherein nearly 100% of the remaining SOX and NO2 are removed. Not all the unit operations in this flowsheet, including SOX removal, were tested in this project. The purified flue gas is sent to a two-stage adsorption system. Particulate adsorbents are used in both stages. In the moisture removal stage, moisture is removed to below 1 ppm. Both the flue gas purification part of the process (removal of SOX and NO2 to below 0.1 ppm levels) and moisture removal to below 1 ppm level were demonstrated at Technology Centre Mongstad (TCM). 86 CO2 Depleted Flue Gas (Dry) 10 CO2 Condensate 18 Adsorption Vacuum Pump CO2 Compressor 12 1 2 3 4 9 13 Cooling 14 16 15 High Water Pressure Hot Water CO2 Product CO2 Depleted Moisture Flue Gas Adsorption 11 Stack To Cooling Tower From Cooling Tower 8 4 5 6 7 Direct Contact Baghouse FGD Cooler Induced To Cooling Tower Draft 17 SCR Fan LP Steam Condensate 2 19 Pulverized Air Coal HP MP LP Boiler Turbine Turbine Turbine BFW 21 Preheater 1 Feedwater 20 3 Heater Coal Ash System Feed Figure 12.2. InnoSepra CO2 capture process schematic and its integration with the power plant The CO2 from dry flue gas is adsorbed in the second adsorption stage. The CO2 adsorption beds contain a sorbent for CO2 adsorption. The process configuration in Figure 12.2 protects the CO2 adsorbent from the flue gas contaminants as well as moisture. The adsorbents used in the InnoSepra process have a significantly higher affinity for CO2 compared to N2, O2, or NO. Nitric oxide (NO) is not removed in the moisture beds or the CO2 adsorption beds, and the NO impurities remain in the dry CO2-depleted stream 10 leaving this section. The CO2 adsorption section consists of multiple beds to enable a continuous process with each bed undergoing the steps of adsorption, rinse with carbon dioxide, heating with evacuation, purging with CO2-depleted product, and cooling in sequence. The adsorbed CO2 is desorbed using a combination of heat and vacuum to produce high-purity CO2 and is sent to the compressor train. 87 The sorbents in the CO2 beds are heated by the regenerant, which in turn is heated by steam from the power plant in a separate heat exchanger (not shown). Regeneration is carried out at temperatures of 100-110 0C. During the heating part of the cycle, CO2 is removed using a vacuum blower and, after cooling to separate the regenerant, is sent to the first stage of the CO2 compressor. The CO2 compressor has four or more stages and compresses the CO2 to about 2,200 psig for pipeline transport. The CO2 product after the 4th compression stage can optionally be condensed at a temperature of about 7oC (45 bara) using cooling water at about 5oC. Liquid CO2 at 7 oC is pumped to a pressure of 2,200 psig. Compared to a standard 7-stage compression process, this configuration saves more than 15% of the energy needed for CO2 compression. The CO2 purity from the InnoSepra process (<10 ppm O2, <5 ppm each of SO2 and H2O) is adequate for direct use in EOR or sequestration applications. Part of the LP steam from the LP/IP crossover, stream 17, is used for heating the regenerant, and condensate, stream 18, is sent to the boiler feed water (BFW) preheater to heat the condensate from the LP turbine. Since the LP stream is at a pressure higher than what is needed for the regeneration of the CO2 beds, it is expanded in a backpressure turbine to produce additional power. Streams 20 and 21 from the BFW preheater are further heated and sent to the PC boiler for steam generation. A small amount of the cooling water stream leaving DCC is purged to prevent salt build-up. The InnoSepra Process, in its commercial embodiment, will use several fixed sorbent beds in parallel to preferentially adsorb CO2 from the flue gas stream, and subsequently liberate the CO2 at high purity via heating the material and subsequently pulling a vacuum to further regenerate the material and produce additional CO2 (Figure 12.3). A dehydration unit is used to remove moisture from the CO2-containing flue gas stream before being introduced to the CO2 adsorption unit. This dehydration unit consists of a direct contact cooler used to reduce the temperature in the gas stream and saturation concentration of water, followed by several parallel fixed beds of desiccant that are used to dehydrate and polish the moisture from the flue gas stream. These beds are regenerated through purging with dehydrated flue gas from which the CO2 has been removed. The feed pressurization is accomplished with a forced draft (FD) fan that takes flue gas from the FGD outlet stream before the stack in a traditional coal-fired power plant. The outlet from this FD fan is to the direct contact cooling (DCC) tower, which circulates water with a small amount of sodium hydroxide present in order to cool the incoming flue gas, lower the moisture content of 88 the incoming gas, and also reduce the acid gas content (SOX and NOX) of the gas below the existing FGD removal levels. After the DCC and flue gas purification, the flue gas is split into several trains of adsorption beds for further moisture removal and CO2 separation. Figure 12.3. The InnoSepra Process integrated with the power plant The InnoSepra Process integrated with the power plant for the techno-economic analysis is shown in Figure 12.3. Because of the scale of the separation required, the adsorption component of the process is designed to run in several independent parallel trains, each of which consists of two drying beds and three CO2 adsorption beds. Due to the size constraints of the largest vessels and the vacuum pumps, the minimum number of trains that need to be employed to capture the CO2 from a 650 MW coal-fired plant is five parallel trains. These five trains produce CO2 at atmospheric pressure and then feed into one compression train that compresses the CO2 product to a pipeline pressure of 2,215 psia (152.7 bar). Heating for the process is provided by two different mechanisms. In the CO2 adsorption beds, 89 the heat transfer medium is the regenerant heated in a heat exchanger by condensing steam extracted at the IP/LP turbine crossover. Because there is significant recoverable energy in the steam between the extraction temperature and pressure and the ultimate temperature and pressure of the steam used to heat the regenerant, a letdown or backpressure turbine is used to recover what would otherwise be wasted recoverable energy. The regenerant from this heating cycle process flows through sorbent beds in the CO2 adsorption system to heat the sorbent and release the CO2. The other heat requirement is for regenerating the moisture removal beds. These moisture removal beds are based on the heatless dryer concept, where humid gas is contacted with a sorbent during the drying step and the bed is subsequently regenerated through purging with a dry gas stream. A small amount of heat may be required to regenerate the dryer beds. The heat required for this step can be fairly low quality. In the case of the InnoSepra process, the heating for the dehydration step can be achieved through heat recovery from the last three stages of the CO2 compression system. CO2 compression is achieved first within each adsorption train via a vacuum pump operating at pressures from 1 bar down to 0.15 bar (14.7-2.2 psia) with an outlet of 1 bar pressure. The atmospheric CO2 stream is then cooled and compressed in a single train of 4 centrifugal compressors with intercoolers and heat integration, and a liquid pump to the required pipeline pressure of 152.7 bar (2,215 psia). An ASPEN Plus simulation for a 650 MWnet SCPC plant with CO2 capture using the InnoSepra process was carried out (Case B12C), and the flowsheet is shown in Figure 12.4. Flue gas (1FLUEGAS) enters the system at 57 oC and 1 bar pressure. Any liquid carryover is removed in a separator (CONDSEP). It flows then (2FDPUR) through a purification bed (PURBED) where NO2 and SO2 are removed. The resulting stream (3FDBLWR) enters the blower (C-101), where the pressure is raised to 1.141 bar (4FDDCC) and enters a Direct Contact Cooler (DCC) to cool down the gas. This saturated stream (5FDDRYR) enters a dryer to generate a dried stream (6FDCO2BD) that goes to the CO2 adsorber beds (ADSCO2BD). The dryer is regenerated with a CO2-depleted stream (7CO2DEPD). The stream coming out of the dryers (8CO2DEPM) is released into the atmosphere. The CO2-rich stream that comes out of the CO2 adsorber beds during the liquid regeneration 90 and vacuum steps (33RGNXAD) goes through a splitter (VPSPLIT) that separates the liquid and vapor phases. The liquid stream (34RGNGRV) flows into a hot liquid tank (RGNTNK). The vacuum pump (P-103) is a two-stage unit that sends the vapor stream to the same hot liquid tank. The liquid regenerant stream (30RGNFD) is fed to a liquid pump P-102 that in turn feeds (31RGNHTR) the vaporizer HX-101 and leaves the unit as a vapor (32RGNADS) that is used to regenerate the CO2 adsorption beds. The CO2-rich stream out of the hot liquid tank (37RGNHEX) is cooled down in a heat exchanger (HX-103) using water at 25oC. The resulting stream (9FDCMP) enters a 4-stage compressor in which the pressure is raised to 55 bar. The compressed stream (10CO2XCM) passes to a guard bed (RGNGB) to remove any traces of the regenerant. The purified stream (11CO2GTL) goes to a heat exchanger (HX-104) cooled by water and, as the temperature drops to 10 oC, the compressed CO2 condenses (11CO2LTP). A liquid pump P-102 raises the liquid CO2 pressure to 155 bar (12CO2LPD). Figure 12.4. ASPEN Plus simulation for a 650 MWnet SCPC plant with CO2 capture (Case B12C). The key results from the Aspen Plus simulation for a 650 MW plant with carbon capture for a 91 feed containing 68.1% N2, 13.5% CO2, and 15.2% water (SCPC plant) at 57oC and 1 bar are shown in Table 12.5. Table 12.5: Key Results from Process Simulation o CO2 product (liquid at 24.7 C and 155 bar): 99.5% purity CO2 recovery: 95.0% Specific power consumption (heating): 1.8 GJ/MT Electrical power consumption: 63.2 MW Power loss due to steam extraction: 28.4 MW Total power loss due to CO2 capture: 91.6 MW - Net power generation: 558.4 MW In order to generate the lost power, a GE LM6000 PG gas turbine operating in the combined cycle mode (Case B12E) was added. Flue gas from the turbine is added to the flue gas from the coal plant. An ASPEN Plus simulation for this configuration was carried out, and the corresponding flowsheet is shown in Figure 12.5. Figure 12.5. ASPEN Plus simulation for a 650 MWnet SCPC plant with CO2 capture (Case B12E) The key results from the Aspen Plus simulation for a 650 MW plant with carbon capture for a 92 feed containing 68.1% N2, 13.5% CO2, and 15.2% water (SCPC plant) at 57oC and 1 bar are shown in Table 12.6. Table 12.6: Capture Process Integrated with a Gas Turbine o CO2 product (liquid at 24.8 C and 155 bar): 99.4% 90.2% (including CO2 recovery: the gas turbine) Specific power consumption 1.8 GJ/MT Electrical power consumption: 66.7 MW Power loss due to steam extraction: 30 MW Total power loss due to CO2 capture: 96.7 MW Net power: 625 MW The total CO2 capture is 90.2% instead of 95% in Case B12C, and the net power loss is 25 MW (less than 4% of the plant's power output). Techno-Economic Analysis The techno-economic analysis summary for Case B12A from the 2019 Baseline Report is shown in Table 12.7. Table 12.7: TEA Summary for Case B12A Total As Spent Capital (TASC) $1,937,000,000 Total Annual Cost (TAC) $311,600,000 LCOE, $/MWh 64.4 The techno-economic analysis summary for Case B12B-Derate (Shell Cansolv) is shown in Table 12.8. Table 12.8: TEA Summary for Case B12B-Derate Total As Spent Capital (TASC) $2,956,000,000 Total Annual Cost (TAC) $420,000,000 93 LCOE with Capture, $/MWh 112.4 Breakeven CO2 Capture Cost, $/MT 46.0 Net Power Generation, MWh 502 The techno-economic analysis summary for Case B12C (InnoSepra Capture) is shown in Table 12.9. Table 12.9: TEA Summary for Case B12C (InnoSepra Capture) Total As Spent Capital (TASC) $2,625,000,000 Total Annual Cost (TAC) $396,000,000 LCOE with Capture, $/MWh 95.4 Breakeven CO2 Capture Cost, $/MT 25.1 CO2 Sales Price (no increase in power cost), $/MT 38.1 Net Power Generation, MWh 558.4 The techno-economic analysis summary for Case B12D (InnoSepra Capture with a gas turbine) is shown in Table 12.10. Table 12.10: TEA Summary for Case B12D (InnoSepra Capture with GE LM7600 PG) Total As Spent Capital (TASC) $2,804,000,000 Total Annual Cost (TAC) $436,700,000 LCOE with Capture, $/MWh 92.8 Breakeven CO2 Capture Cost, $/MT 36.9 CO2 Sales Price (no increase in power cost), $/MT 39.4 Net Power Generation, MWh 632 The results from the evaluation of relevant TEA cases are summarized in Table 12.11. 94 Table 12.11. Summary of various TEA cases Case B12A Case Case Case Case B12B-Derate B12C-Inno B12D-Inno B12E-Inno TASC, $MM 1,937 2,956 2,625 2,804 2,884 TAC, $MM 311.6 420.0 396.0 436.7 443.7 LCOEa, $/MWH 64.4 112.4 95.4 92.8 93.8 Breakeven CO2 Capture - 46.0 25.1 36.9 37.4 Costb, $/MT CO2 Sales Price, no - - 38.1 39.4 40.7 increase in power costc, $/MT Net Power, MWh 650 502 558.4 632.0 626.0 Notes: a With or without capture b No T&S c The LCOE in these cases is the same as Case B12A, $64.4/MWh. The CO2 sale prices for cases B12C, B12D, and B12E for no increase in power cost are very close (the capture plant pays for the power needed for capture). The breakeven CO2 capture cost for the InnoSepra Process, Case B12C, is about 45% lower than the breakeven capture cost for Shell Cansolv, Case B12B-Derate. The power loss due to CO2 for the InnoSepra Process is 91.6 MW (14.1% of plant output) compared to a power loss of 148 MW (23% of plant output) for Shell Cansolv. The loss in power output can be made up for by adding a gas turbine to the capture process. For both gas turbine cases (Case B12D with no CO2 capture from the gas turbine, and Case B12E with CO2 capture from the gas turbine), the breakeven 95 CO2 capture cost does increase by $11-12/tonne. However, when the CO2 sales price for no increase in power cost (the loss in revenue due to reduced power generation is made up by increasing the CO2 sales price) is considered, all three cases (B12C-Inno, B12D-Inno, and B12E- Inno) are very close to each other ($38-$41/tonne). The highest CO2 capture (90% from both the power plant and gas turbine), coupled with a very small loss in power output (3.7%), is achieved for Case B12E-Inno. Task 13 – State Point Data Table The objectives of this task were to finalize and submit a complete State Point Data Table in accordance with FOA Appendix C. The State Point Data Table, based on lab and field tests, is provided in Table 13.1. The values in this Table have either been provided by the vendor or measured in the lab/field. Table 13.1: State Point Data Table Measured / Projected Units Estimated Values Performance Sorbent True Density @STP kg/m 3 1,990 1,990 3 Bulk Density kg/m 690 660-680 Average Particle Diameter mm 1.6 1.4-1.8 Particle Void Fraction m 3/m 3 0.45 0.45 2 3 Packing Density m /m 1.79E+08 1.79E+08 Solid Heat Capacity @STP kJ/kg-K 0.96 0.96 Crush Strength - 2.9 2.9 Attrition Index 0.10% <0.10% Adsorption Pressure bar 1.15 1.1-1.15 o Temperature C 30 25-30 Equilibrium Loading gmol CO2/kg 3.25 3.1-3.3 Heat of Adsorption kJ/mol CO2 40 40 Desorption Pressure bar 0.25 0.20-0.25 o Temperature C 100 100-105 Equilibrium Loading gmol CO2/kg 1.1 1.0 Heat of Desorption kJ/mol CO2 40 38-40 96 Task 14 – Technology Gap Analysis (TGA) The objectives of this task were to prepare and submit a complete TGA in accordance with FOA Appendix E. Regenerant CO2 LP STM In Depleted FG #5 HE Cond Out #6 Adsorption #3 CO2 Adsorption CO2 to Compression Regenerant/CO2 Mixture #4 Condenser #2 Moisture Adsorption #1 FG Purification From DCC Figure 14.1. Flowsheet for Technology Gap Analysis The flowsheet for Technology Gap Analysis is shown in Figure 14.1. The new unit operations for the InnoSepra Process with the end of project TRLs include: • Flue Gas Purification – TRL 7. This unit removes impurities such as SOX and NO2 from the flue gas prior to the CO2 capture step. Based on testing during this project and prior testing, this step does not have a significant technical risk. However, the scale-up is needed to address potential flow distribution issues. Applicable to all CO2 capture technologies. • Moisture Removal – TRL 6/7. This unit removes moisture from the flue gas prior to CO2 adsorption in Unit #3. The moisture removal prior to CO2 adsorption increases the CO2 97 capacity for the sorbent used in this process. Based on testing during the project, this step is not a significant technical risk. However, the scale-up is needed to address potential flow distribution issues. Applicable to CO2 capture technologies that do not involve steam regeneration. • CO2 Sorbent Regeneration – TRL 4. This unit adsorbs CO2 from dry flue gas and produces a dry CO2 product. Complete pilot-scale testing of this step could not be carried out during this project. All the steps in this unit operation were tested at the lab scale following the completion of TCM testing. This step needs the most attention in future work. It is very promising for reducing parasitic power consumption. However, only parts of this technology have been demonstrated at the pilot scale. 98 Task 15 – Environmental Health and Safety Risk Assessment The objectives of this task were to prepare and submit an EH&S assessment in accordance with FOA Appendix F. The key process streams from the InnoSepra Process were evaluated for their potential EH&S impact. The waste streams from the InnoSepra Process (Figure 15.1) include: Vent Gas 10M with NaOH Regenerant Direct Flue Gas Moisture Contact Purification Removal System Flue Gas Cooler CO2 Recovery CO2 Product System Purge Water Recycle Water Dessicant CO2 Sorbent Waste Waste Figure 15.1. Key waste streams • Cooling water from the Direct Contact Cooler at a pH of about 10. This stream is common to all CO2 capture processes and can be disposed of in sanitary sewers. • The flue gas purification unit removes SOX and NO2 by reaction. Both the unused and spent sorbents from the flue gas purification unit are non-toxic and can be disposed of in non- hazardous landfills (replaced 3 times a year). • The moisture removal sorbent undergoes hydrothermal degradation over time and would need to be replaced about every five years. Both the unused and spent sorbents from the moisture removal unit are non-toxic and can be disposed of in non-hazardous landfills. • The CO2 removal sorbent undergoes thermal degradation over time and would need to be replaced about every five years. Both the unused and spent sorbents from the CO2 capture unit are non-toxic and can be disposed of in non-hazardous landfills. 99 Lessons Learned from TCM Testing • Need to finalize any regulatory issues, such as approval for chemicals used in the process, before testing. • Project delays pushed the testing to Q1-24, leading to significant freezing issues that were not accounted for in the equipment design. In hindsight, the equipment should have been winterized. • A prior relationship with the project team is very helpful in executing various project tasks. • Lower TRL unit operations are much more challenging to commission and test at TCM. This is due to field updates and modifications needed to the equipment, and delays in getting replacement parts and supplies. Additional spares could have been ordered to mitigate some of the delays. • Despite all the issues, the test results are very valuable for not only adsorption-based capture technologies but also for solvent and membrane-based capture technologies. The flue gas purification part of the process is very valuable for the solvent-based capture processes as it removes nearly 100% of SOX and NO2, and significant amounts of aerosols. Removal of these impurities prior to solvent-based capture would significantly reduce solvent degradation and environmental emissions. The flue gas drying part of the process can benefit both the membrane-based capture processes and the capture processes based on freezing. The CO2 compression cycle developed during this project can benefit all capture technologies. Summary The InnoSepra CO2 capture technology has the potential for a significant reduction in the CO2 capture cost for the power plant and industrial flue gases. The process can provide both very high recovery (90-95%) and high purity (>98%) CO2 with physical sorbents while meeting the EOR/sequestration product specifications. The process has the potential to reduce the thermal parasitic power by more than 65%, and the capital required by about 45% leading to about 45% reduction in the CO2 capture cost for the coal-based power plant flue gas compared to amine-based capture processes. After demonstration at a bigger scale and further process optimization, the process has the potential for a capture cost of around $40/MT with no increase in LCOE and with very little loss in power output. 100 Summary of Capture Process Co-benefits While InnoSepra’s CO2 capture technology can provide a significant reduction in the CO2 capture cost, it also provides significant co-benefits for all the CO2 capture technologies. InnoSepra’s flue gas purification technology provides a low-cost option for the removal of NO2 and SOX to sub-ppm levels and a significant reduction in aerosol emissions, and is applicable to all solvent-based capture technologies. It can significantly reduce emissions from solvent-based capture processes. InnoSepra’s CO2 compression technology can reduce the cost of CO2 compression to pipeline pressures by about 20%. It is applicable to both solvent-based capture (3.4 bara CO2 pressure) as well as membrane and sorbent-based capture (~1 bara CO2 pressure). InnoSepra’s flue gas drying technology can be used with other capture technologies, such as membranes and capture by freezing, to reduce capture costs. It can also be upgraded for regenerative removal of NO2, SO2, and Hg, potentially eliminating the need for new FGDs and SCRs for industrial capture. 101 Bibliography 1. DOE/NETL, “Cost and Performance Baseline for Fossil Energy Plants Volume 1a: Bituminous Coal (PC) and Natural Gas to Electricity”, Revision 3, DOE/NETL-2015/1723, 2015 2. DOE/NETL, “Development of CO2 Purification Unit and Compression System Spreadsheet Tool”, DOE/NETL-4001/080415, August 2015 3. DOE/NETL, “Cost and Performance Baseline for Fossil Energy Plants Volume 1: Bituminous Coal and Natural Gas to Electricity”, NETL-PUB-22638, 2019 4. Jain, Ravi – “Transformational Sorbent-Based Processes for a Substantial Reduction in the Cost of CO2 Capture”, August 2019 5. https://netl.doe.gov/sites/default/files/netl-file/R-Jain-InnoSepra-Cost-Reduction-Capture.pdf 6. Jain, Ravi, “Transformational Sorbent-Based Process for a Substantial Reduction in the Cost of CO2 Capture”, 2019 Carbon Management Technology Conference, Houston, July 2019 7. Jain, Ravi, “Transformational Sorbent-Based Process for a Substantial Reduction in the Cost of CO2 Capture”, 2022 Carbon Management Project Review Meeting, Pittsburgh, August 2022 8. Jain, Ravi, and N. Lemcoff, “Transformational Sorbent-Based Process for a Substantial Reduction in the Cost of CO2 Capture”, 2023 FECM/NETL Carbon Management Research Project Review Meeting, Pittsburgh, August/September 2023 9. Jain, Ravi, “Transformational Sorbent-Based Process for a Substantial Reduction in the Cost of CO2 Capture”, 2024 FECM/NETL Carbon Management Research Project Review Meeting, Pittsburgh, August 2024 102