Power-to-Syngas: A Parareal Optimal Control Approach

Andrea Maggi, Dominik Garmatter, Sebastian Säger, Martin Stoll, Kai SundmacherFrontiers in Energy Research 2021 · DOI

Abstract

A chemical plant layout for the production of syngas from renewable power, $H_2O$, and biogas is presented to ensure a steady productivity of syngas with a constant $H_2$-to-$CO$ ratio under time-dependent electricity provision. An electrolyzer supplies $H_2$ to the reverse water-gas shift (RWGS) reactor. The system compensates for a drop in electricity supply by gradually operating a tri-reforming (TRI) reactor, fed with pure $O_2$ directly from the electrolyzer or from an intermediate generic buffering device. After the introduction of modeling assumptions and governing equations, suitable reactor parameters are identified. Finally, two optimal control problems are investigated, where computationally expensive model evaluations are lifted via parareal and necessary objective derivatives are calculated via the continuous adjoint method. For the first time, modeling, simulation, and optimal control are applied to a combination of the reverse water-gas shift and tri-reforming reactor, exploring a promising pathway in the conversion of renewable power into chemicals.

Introduction: The Challenge of Intermittent Renewables

The transition to a carbon-neutral economy relies heavily on "Power-to-X" technologies—converting renewable electricity into chemicals or fuels. Syngas, a mixture of $H_2$ and $CO$, is a vital intermediate for producing liquid fuels via Fischer-Tropsch synthesis. However, a fundamental mismatch exists: chemical reactors perform best under steady-state conditions, while renewable energy sources like wind and solar are notoriously intermittent. Existing solutions often rely on massive hydrogen storage, which is energy-intensive and expensive. This research explores an alternative: a hybrid plant design that uses secondary reactors to "fill the gaps" when power drops, managed by advanced mathematical control strategies.

Proposed Plant Layout: A Hybrid Reactor System

The proposed hybrid plant layout consists of an electrolyzer (EL), a Reverse Water-Gas Shift (RWGS) reactor, and a Tri-Reforming (TRI) reactor. When renewable power is abundant, the electrolyzer provides $H_2$ to the RWGS reactor. When power drops, the system ramps up the TRI reactor, which uses biogas and oxygen (stored from the electrolysis process) to maintain the syngas flow.

Overall plant layout: the RWGS and TRI reactors produce raw syngas, where H2 and O2 are supplied via EL.

By treating the $O_2$ produced during electrolysis as a resource rather than a waste product, the system drives exothermic reactions in the TRI reactor, helping to balance the plant's energy needs. This dynamic flexibility offers a promising pathway for Power-to-X, providing a more adaptable alternative to simple hydrogen buffering.

Computational Efficiency via Parareal Control

To manage this complex system, the researchers formulated an Optimal Control Problem (OCP) to minimize deviations from the target syngas ratio ($H_2/CO = 2$) and flow rate. Solving these problems is computationally expensive because it requires simulating complex chemical kinetics over time.

To address this, the authors utilized the Parareal algorithm, a parallel-in-time method that breaks the time domain into smaller chunks. It uses a "coarse" (fast but less accurate) model to coordinate between chunks and a "fine" (slow but accurate) model to solve them in parallel across multiple CPU cores. Additionally, the continuous adjoint method was used to calculate the necessary mathematical gradients for the optimizer, ensuring the control system could find the most efficient settings.

Key Results and Performance

The study demonstrated significant improvements in both computational speed and plant stability:

  • Speedup: The Parareal approach achieved a speedup of 2.57x for standard academic problems and 1.70x for realistic, fully time-dependent plant simulations compared to traditional sequential methods.
  • Steady Productivity: The controller successfully maintained a constant $H_2/CO$ ratio of 2.0 and a steady flow rate of 111.44 mol/s, even during significant simulated drops in electricity supply.
  • Validation: The results confirmed that the electrolyzer is the dominant power consumer, justifying the focus on managing its fluctuations to stabilize the rest of the plant.

Conclusion and Future Directions

While this research proves the technical feasibility of a hybrid RWGS/TRI layout, several areas for future work remain. The current model neglects the internal dynamics of the electrolyzer and $O_2$ buffer, treating them as quasi-steady-state. Furthermore, the TRI reactor experiences high temperature peaks that require more sophisticated $O_2$-dosing strategies to prevent catalyst damage. Finally, a thorough economic analysis is needed to compare the commercial viability of this hybrid layout against traditional hydrogen buffering solutions. Despite these limitations, the integration of modeling, simulation, and parareal optimal control marks a significant step forward in converting intermittent renewable power into stable chemical production.

Summary updated

Abstract

A chemical plant layout for the production of syngas from renewable power, H 2 O and biogas, is presented to ensure a steady productivity of syngas with a constant H 2 -to-CO ratio under time-dependent electricity provision. An electrolyzer supplies H 2 to the reverse water-gas shift reactor. The system compensates for a drop in electricity supply by gradually operating a tri-reforming reactor, fed with pure O 2 directly from the electrolyzer or from an intermediate generic buffering device. After the introduction of modeling assumptions and governing equations, suitable reactor parameters are identified. Finally, two optimal control problems are investigated, where computationally expensive model evaluations are lifted via parareal and necessary objective derivatives are calculated via the continuous adjoint method. For the first time, modeling, simulation, and optimal control are applied to a combination of the reverse water-gas shift and tri-reforming reactor, exploring a promising pathway in the conversion of renewable power into chemicals.
BibTeX
@inproceedings{maggi2021power,
  author = {Andrea Maggi and Dominik Garmatter and Sebastian Säger and Martin Stoll and Kai Sundmacher},
  title = {{Power-to-Syngas: A Parareal Optimal Control Approach}},
  booktitle = {Frontiers in Energy Research},
  year = {2021},
  doi = {10.3389/fenrg.2021.720489},
}
Figure