Model-based optimization and experimental investigation of CO2 methanation

Abstract

Today, renewable energy systems are perceived by the scientific communities and the public opinion as crucial tools for meeting future energy demands, ensuring the security energy supplies, and combating global warming. The intermittent nature of renewable sources, nevertheless, makes energy storage a key component in any renewable energy system. In this regard, power-to-gas concept (P2G) offers an efficient means to store excess renewable energy in large scale and for periods ranging from hours to months. Given that the infrastructure for transporting and storing natural gas is already established in many countries, converting renewable electricity into H2 and then into Synthetic Natural Gas (SNG) via CO2 methanation appears to be an attractive P2G process route. The recycling of CO2 is another main advantage of this route. This thesis investigates the CO2 methanation in the context of P2G application on various hierarchical levels: chemical kinetics, reactor, and overall process. In the first part of this work, a model-based optimal reactor design is developed for the production of high quality SNG that can be ready for grid injection following a downstream drying step. The three-level reactor design procedure of Peschel1, which is based originally on the flux-oriented Elementary Process Function (EPF) methodology2, is implemented. On level 1, predefined apparatuses are not considered. Instead, the reactor is represented by a matter element in the thermodynamic state space. The matter element is manipulated over the entire reaction time by unlimited mass and heat fluxes so that optimal profiles of the state variables (e.g. T, xi) are achieved. The effect of optimal dosing/removal of certain components and optimal cooling strategies on the reactor productivity is systematically investigated. Based on the results of Level 1, it is found that continuous water removal combined with active cooling is the most promising approach for enhancing the reactor performance. Consequently, on Level 2, two reactor configurations are considered and compared on the basis of their space-time yield: (1) a cascade of polytropic multi-tubular packed bed reactor with intermediate Word Template by Friedman & Morgan 2014 condensation steps and (2) a polytropic hydrophilic membrane reactor. For each case scenario, optimal geometric designs and operating conditions are determined such as the inlet conditions, temperature profiles, reactor dimensions. It is concluded that a cascade of three condensation steps and four externally cooled multitubular reactor is the best configuration for maximizing space-time yield. On level 3, the optimal temperature profiles are technically realized. Furthermore, the proposed configuration is compared with the traditionally adopted cascade of adiabatic reactors with intercooling steps or a single polytropic reactor. The results show that our proposed reactor configuration performs significantly better in terms of productivity. Moreover, using our configuration, it is possible to have good temperature control inside the reactors and prevent the formation of hotspots, thus lengthening the life time of the catalyst. The second part of this work deals with the optimal design and operation of the whole methanation process with main emphasis on the reactor design. The considered process design comprise of a reaction section and a drying section. The reaction section is a cascade of reactors and condensers, while the drying section is represented by a glycol dehydration unit. While the reactors are rigorously modeled, all other process units (compressors, pumps, heat exchangers, and absorber) are described using short-cut models. Since the reactors and the various process units are simultaneously optimized for reduced operating and investment costs, the optimization problem includes cost models as well. It is shown that the cost optimal reaction section is a cascade of three multitubular reactors with two intermediate condensation steps for water removal. Furthermore, the results clearly demonstrate the significant importance of accounting for the interaction between the reaction section and the other process units when determining the optimal reactor design and process operating conditions. In the last part of this work, a comprehensive mechanistic kinetic model that best described CO2 methanation reaction over a commercial nickel-based catalyst is developed. It is proposed that CO2 methanation occurs through a series of reactions, namely the r-WGS and the CO methanation. As such, the derived model is comprised of Word Template by Friedman & Morgan 2014 the rate expressions of these two reactions. Data collected over a wide range of industrial relevant conditions (temperature, pressure, and feed content) is used in discriminating between rival models and in estimating the parameters. The parameters are estimated by minimizing the determinant of the matrix of the sum of squares and the cross-products of the residuals. Furthermore, the method of Vajda3,4, which is based on eigen-decomposition of the Fisher Information Matrix, is used to investigate the quantitative identifiability of the parameters of the most adequate model. Based on this investigation, it is shown that out of the twelve model parameters, eight are uniquely identifiable. Also, the predictive ability of the model is validated using a different set of data that wasn not used in the estimation.