Search

In order to increase the typically low equilibrium CO2 conversion to methanol using commercially proven technology, the addition of two intermediate condensation units between reaction steps is evaluated in this work. Detailed process simulations with heat integration and techno-economic analyses of methanol synthesis from green H2 and captured CO2 are presented here, comparing the proposed process with condensation steps with the conventional approach. In the new process, a CO2 single-pass conversion of 53.9% was achieved, which is significantly higher than the conversion of the conventional process (28.5%) and its equilibrium conversion (30.4%). Consequently, the total recycle stream flow was halved, which reduced reactant losses in the purge stream and the compression work of the recycle streams, lowering operating costs by 4.8% (61.2 M€·a−1). In spite of the additional number of heat exchangers and flash drums related to the intermediate condensation units, the fixed investment costs of the improved process decreased by 22.7% (94.5 M€). This was a consequence of the increased reaction rates and lower recycle flows, reducing the required size of the main equipment. Therefore, intermediate condensation steps are beneficial for methanol synthesis from H2/CO2, significantly boosting CO2 single-pass conversion, which consequently reduces both the investment and operating costs.

Hybrid kinetic models represent a promising alternative to describe and evaluate the effect of multiple variables in the performance of complex chemical processes, since they combine system knowledge and extrapolability of the (semi-)mechanistic models in a wide range of reaction conditions with the adaptability and fast convergence of data-based approaches (e.g., artificial neural networks—ANNs). For the first time, a hybrid kinetic model for the direct DME synthesis was developed consisting of a reactor model, i.e., balance equations, and an ANN for the reaction kinetics. The accuracy, computational time, interpolation and extrapolation ability of the new hybrid model were compared to those of a lumped and a data-based model with the same validity range, using both simulations and experiments. The convergence of parameter estimation and simulations with the hybrid model is much faster than with the lumped model, and the predictions show a greater degree of accuracy within the models’ validity range. A satisfactory dimension and range extrapolation was reached when the extrapolated variable was included in the knowledge module of the model. This feature is particularly dependent on the network architecture and phenomena covered by the underlying model, and less on the experimental conditions evaluated during model development.