Dynamics and Control of R2Chem Processes & Reactors (R2Chem)

Utilizing volatile renewable energy sources (solar, wind, biomass) in chemical production systems requires not only advanced methods for optimal system design, but also deep understanding of their dynamic operation and the establishment of powerful optimal control strategies. In this project, such investigations are performed from the plant level down to the phase and even molecular level.

When converting renewable energies into chemical products, fluctuating price and quality characteristics inevitably lead to the question of suitable buffer strategies. Although, these buffers will increase the investment costs, they also increase the overall operation time in such scenarios. Consequently, there exists a trade-off between investment and operating costs, influenced by the dynamic capabilities of the process itself and its corresponding process units. In this context, Fig. 1 shows a Power-to-Gas (PtG) process network as one promising technology for future energy conversion and storage. To equalize different production loads of the individual process units (biogas plant, electrolysis, methanation) the correct location and dimension of internal process buffers need to be determined. Therefore, individual process units as well as the entire process are exposed to load changes. Each load change is classified in two phases - the transition and the production phase, which are essential to identify the optimal buffer placement and dimension. The use of modern optimization-based control approaches is therefore of major importance and under ongoing investigation [1].

Fig. 1: Power-to-Gas process network and possible buffers that allow for improved dynamic operation.

Dynamic perturbations caused by the use of volatile renewable resources are not only critical for a whole production process system, but also for certain process units. An important example is the fixed-bed methanation reactor (see Fig. 1), which is very sensitive to fluctuating feed streams due to the strongly exothermic reaction of hydrogen and carbon dioxide. Thus, controlling its temperature profile is an important task. In literature, this reactor type was mostly analyzed under steady-state conditions due to the fact that today most of the catalytic systems in industry operate at steady-state conditions. The PSE group developed a dynamic optimization approach to identify control trajectories for the time-optimal reactor start-up, thereby avoiding distinct hot spot formation within the reactor, based on a dynamic two-dimensional reactor model (Fig. 2) and a three-step kinetic scheme of the underlying Sabatier reaction.

Fig. 2: Multi-level model used to analyze dynamic behavior and controllability of reactive systems.

Dynamic optimization of the 2D methanation reactor model is computationally challenging due to spatially distributed states (concentrations, temperature) and nonlinear equations (transport coefficients, reaction kinetics). Surrogate models with sufficient accuracy represent a remedy to this problem. Featuring a lower number of states, model order reduction (MOR) generates considerably less complex models and leads to faster model evaluations. Especially for nonlinear systems, snapshot-based MOR techniques are considered to be one of the most promising methods. In collaboration with the CSC group (Prof. Peter Benner), we applied proper orthogonal decomposition together with the discrete empirical interpolation method (POD-DEIM) to the dynamic, two-dimensional reactor model for CO2 methanation [3-4]. It was shown that the reduced order model (ROM) is accurate and, furthermore, the solution is accelerated at least by one order of magnitude compared to the full order model.

Additionally to the aforementioned missing knowledge about reactor dynamics, the impact of dynamically changing catalyst conditions has also not been investigated sufficiently yet. However, research shows that the structure of the catalyst, and therefore the activity and stability, strongly depends on the reaction conditions and can change for example by phase transitions or sintering. Additionally, accumulation terms for mass and energy influence the temporal evolution of the reaction systems. In order to handle changing reaction conditions, for instance induced by the volatile nature of feedstocks coming from renewables, it is necessary to obtain fundamental understanding of the of the underlying physio-chemical processes at all time and length scales, involving the reactor scale, the particle scale and the atomic scale as shown in Fig. 3. These processes are studied experimentally and the gained knowledge about the catalyst dynamics is used to extend the aforementioned reactor model. This pursued multi-scale approach allows for the derivation of a new catalyst-reactor concept, which is optimally suited for a more flexible reactor operation [5]. 

Fig. 3: Time and length scales involved in dynamic catalytic fixed-bed reactor operation.


[1] A. Himmel, S. Sager, K. Sundmacher, Time-optimal set point transition for nonlinear systems. Automatica, (2018), (submitted)

[2] J. Bremer, K.H.G. Rätze, and K. Sundmacher, CO2Methanation: Optimal Start-Up Control of a Fixed-Bed Reactor for Power-To-Gas Applications. AIChE J. (2017), 63, 23–31.

[3] J. Bremer, P. Goyal, L. Feng, P. Benner, and K. Sundmacher, Nonlinear Model Order Reduction for Catalytic Tubular Reactors, In: Computer Aided Chemical Engineering (2016), 38,  2373-2378.

[4] J. Bremer, P. Goyal, L. Feng, P. Benner, and K. Sundmacher, POD-DEIM for Efficient Reduction of a Dynamic 2D Catalytic Reactor Model. Computers & Chemical Engineering (2017), 106, 777-784. 

[5] R.T. Zimmermann, J. Bremer and K. Sundmacher, Improved Catalyst Concept for Thermal Sensitivity Reduction, Performance Improvement and Flexibility Enhancement of Fixed Bed Reactors for CO2Methanation (in preparation)

Go to Editor View