Decarbonizing industry by means of thermochemical conversion of biomass and CO2 capture & transport

Abstract

[EN] The Paris Agreement, conceived to combat climate change, commits member states to limit global warming to 2 °C above pre-industrial levels, requiring an 80- 95% reduction in greenhouse gas emissions from 1990 levels by 2050. Industrial activities, responsible for a quarter of global emissions, must undergo decarbonization for these targets to be met. Eight categories of industrial decarbonization options are identified: (1) energy-efficiency improvements, enhancing energy efficiency can cut fuel consumption by 15-20%; (2) renewables, transitioning to renewable energy sources; (3) energy storage, addressing variability in renewable energy production through energy storage systems; (4) electrification of heat, shifting from fossil fuel-generated heat to electric-powered systems; (5) hydrogen usage, as a feedstock and substituting fossil fuels with zero-carbon hydrogen in industrial processes in process when high temperature is needed and electrification of heat is not efficient; (6) biomass usage, utilizing sustainably produced biomass as a fuel or feedstock; (7) carbon capture, collecting and storing or utilizing CO2 emissions from industrial processes; and (8) demand-side measures, reducing industrial product demand to lower production and emissions. The thesis focuses on biomass usage (6) and carbon capture (7). Sustainably produced biomass can be used in place of some fuels and feedstocks such as coal or natural gas due to, when sustainably managed, biomass is considered carbonneutral. Carbon Capture, Utilization, and Storage (CCUS) technologies involve capturing CO2 emissions from large sources and either storing it underground or utilizing it in various processes. CCUS plays a crucial role in achieving emissions reduction goals. While recognized as vital by international bodies, there are debates about the viability of CCUS, with some publications questioning its effectiveness. Despite scepticism, CCUS is acknowledged as a key tool for industrial decarbonization. The EU emphasizes CCUS deployment in energy-intensive industries and bioenergy-based plants. Consequently, Bioenergy with Carbon Capture and Utilization (BECCU) and Bioenergy with Carbon Capture and Storage (BECCS) are addressed. BECCU and BECCS processes differ in their CO2 emission balance, with BECCU considered neutral and BECCS having potential negative emissions. The thesis delves into the CIUDEN Foundation, a Spanish public foundation focused on energy research and development, particularly in the ‘three D´s’: decarbonization, digitalization, and decentralization. CIUDEN's Experimental Platform, initiated in 2009, was conceived to develop oxy-combustion and CO2 capture from coal. Nevertheless, considering that coal is no longer an option in Spain, the focus of the thesis is to propose the production of biogenic CO2 from oxycombustion of biomass as a sustainable process for CIUDEN's Experimental Platform; this biogenic CO2 could either be utilized onsite or transported in liquid phase, with both scenarios mainly considering its ultimate purpose in producing e-fuels. To achieve this goal, the thesis is divided into the following sections and steps: • Section #1 - Oxy-combustion of biomass: step #1.1 to conduct a screening process to determine the optimal approach for oxycombustion of biomass, considering existing units; step #1.2 to validate a data reconciliation model for oxy-combustion using experimental data from oxy-coal processes with varying oxygen concentrations and step #1.3, illustrating a case study on biogenic CO2 production at CIUDEN, serving as the target operating point. • Section #2 - CO2 capture: step #2.1 to validate a data reconciliation model for the warm section of the compression and purification unit (CPU) and a simulation model for the cold section. Experimental data from the existing CPU with flue gases from an oxy-mode coal-burning boiler are used for the validation. Simultaneously, the simulation model for the cold section is focused on the thermodynamic model behaviour when impure CO2 is considered, being also validated with experimental data from the CO2 compression and purification unit (CPU); step #2.2: execute the previous selected case study in both the reconciliation model (warm section) and the simulation model (cold section) to characterize the produced CO2 stream. • Section #3 - Impure CO2 transport: step #3.1, to develop a steadystate flow model for impure CO2, validated with experimental data obtained from the transport rig using a real CO2 captured stream containing different contaminants such us moisture, SO2, and oxygen. Starting the summary with the first section, as a conclusion of a screening process done in the thesis, the Circulating Fluidized Bed (CFB) boiler was selected as the best option to achieve the objective. Fluidized bed boilers operate based on the combustion of fuel within a bed of sand, limestone, and combustion residues, exhibiting fluid-like behaviour. The retrofitting of a pulverized coal (PC) boiler introduces complexities, requiring a duplicate fuel handling system, additional auxiliary equipment, and a clean and uniform biomass fuel supply. In contrast, CFB boilers offer advantages over PC boilers: (1) higher residence times, CFB boilers have longer residence times due to cyclone recirculation of solids, providing flexibility in fuel quality; (2) temperature control, the loop seal allows the regulation of bed temperature, contributing to the flexibility in the operation; (3) lower operation temperatures, CFB boilers operate at lower temperatures, reducing NOx formation; (4) no milling process required, CFB boilers are versatile with various fuels, including those in less ideal conditions, requiring no milling process; (5) safety operation, CFB boilers are safer for fuels with high volatile contents, minimizing ignition risks in the fuel feeding system; (6) de-SOx and de-NOx in the furnace, that is directly related with the higher residence times previously explained and the option of injecting limestone and ammonia for lower SO2 or NOx concentrations respectively. The primary focus of the CIUDEN´s CFB modifications lies in the latter part of the feeding system, downstream of the fuel transportation belts. In addressing this, a significant challenge arises due to the lower volume density of biomass compared to coal, resulting in a reduction in the boiler's thermal power below the nominal 15 MWth. Apart from that, certain restrictions and minimum values must be adhered to, including a minimum primary comburent of 9,100 kg/h, a minimum global O2 concentration of 21%v in the comburent streams, and a minimum ascensional velocity of 3.4 m/s. These specified values dictate the minimum biomass inlet necessary for an excess of oxygen in the flue gases, thereby determining the minimal thermal power achievable. But there are more challenges to be studied: issues like chlorine content causing corrosion and potassium content affecting ash melting point are considered. The focus then shifts to the use of Sorbacal SPS® as a sorbent for SOx control in dry sorbent injection (DSI) system. The composition and advantages of Sorbacal SPS® are discussed, highlighting its efficiency in removing acidic gas components. The first section concludes with an analysis of ash behaviour during biomass combustion, considering ash fusion temperatures. Binary systems (CaO-SiO2, K2OSiO2, MgO-SiO2, Na2O-SiO2) are examined, with a focus on potential issues arising from certain compositions. Ternary systems (SiO2-K2O-MgO, SiO2-Na2O-CaO) are introduced for a theoretical analysis, emphasizing the need for experimental validation. In the second section, focused on the CO2 capture, extensive testing involving steady-state and dynamic tests are described on the integration of the CFB boiler and CPU using coal. The warm section of the CPU was deemed suitable for handling biomass flue gases considering minor modifications, because four critical aspects are identified. Firstly, the air-ingress (or infiltration air), is analysed through a comparison of the 'base case' and the 'objective case.' The results indicated that maintaining an identical level of infiltration observed in the baseline scenario with coal is feasible for biomass. Secondly, the desulfurization unit condensing water posed a challenge due to the 3.6 times higher quantity of water in biomass flue gases compared to coal. Existing heat exchangers were found insufficient; this situation is repeated with the third bullet, the drying process at low pressure faced challenges, where the existing heat exchangers are incapable of condensing all the water. Lastly, the behaviour of NOx during compression and pre-cooling revealed a transformation of NO to NO2 under increased pressure and decreased temperature, suggesting a need for model modification when Peng-Robinson is selected as thermodynamical model for the unit. Additional findings include the successful operation of the CFB boiler in conjunction with the CPU, demonstrating a near-zero emission plant. The control philosophy of the integrated plant was tested in various scenarios, including stationary and transient periods, malfunction periods, and emergency situations, showcasing the CPU's ability to follow the boiler during disturbances. Finally, the third section of the thesis, i.e. the CO2 transport grid section. A one-dimensional model addressing CO2 transport with impurities in the dense phase was implemented in ChemCAD® and validated using experimental data from a semiindustrial transport rig operating between 10-31 ºC and 80-110 bar(a). Two thermodynamic models, Peng-Robinson (PR) and Soave-Redlich-Kwong (SRK), are employed for validation, testing low concentrations of impurities (O2, SO2, and moisture) to prevent rig damage. Results showed that 100 ppm of SO2 increased pressure drop, while 500 ppm of oxygen decreased it. Although the chosen models correctly predicted the impure stream's experimental behaviour, they did not align with absolute or relative results for oxygen. A specific study on oxygen density prediction models revealed that Benedict-Webb-Rubin (BWR) accurately matched experimental results at 158 barg and almost 31 ºC, considering an impurity concentration of 5.1 in synthetic air—a notably higher concentration compared to previous cases. This comprehensive analysis highlights the successful implementation of the CO2 transport model while emphasizing the importance of considering impurities and selecting appropriate thermodynamic models for accurate predictions.