(459c) A Plantwide Control Structure for Improved Flexible Production of Green Hydrogen

Rising environmental awareness has recently revived interest in using Hydrogen as an energy vector for renewable electricity through water electrolysis. However, the question of how to operate the electrolysis process safely and efficiently in the face of renewable power uncertainty is still largely unanswered [1]. In the case of reduced power, the Hydrogen-to-Oxygen ratio (HTO) at the anode is known to increase, posing a security risk which can force the plant to be shut down [2]. The effect of increased HTO is further exacerbated in the case of alkaline electrolysis through high values of the pressure and the recirculated lye flowrate, both of which are desired from an operational standpoint [3]. Researchers have suggested different strategies to overcome these limitations of flexible operation, for example through use of Model Predictive Control (MPC) [4], or operational strategies that involve scheduled switching between operating modes [1]. In our work we propose a simple control strategy which extends the feasible operating range using standard advanced control elements. The control configuration is found using systematic procedures for plantwide control and tested against historical weather data.

The plant considered in this work consists of a 2 MW rated electrolyzer together with relevant auxiliary units and is directly supplied with energy through a 2 MW rated wind turbine. The model has recently been presented by Cammann et al. [5][1] and is analyzed according to the “Top-Down, Bottum-Up” procedure of Skogestad for control structure design [6]. Firstly, the operational objective is defined as maximizing the momentary Hydrogen production considering the power input of the wind turbine as unknown disturbance. Operational constraints are set according to the physical limits of the equipment, as well as safety limits for the plant. Importantly, the HTO constraint is explicitly considered. The model is subsequently optimized for realizations of the power input varying from 15 kW to 2 MW using available degrees of freedom. Control pairings are proposed for each of the regions in which either the manipulated or control variables saturate, together with selectors to switch between the relevant regions. Lastly, the performance of the control structure is evaluated by simulating it with wind power data at the wind park location of Hollands Kust Zuid [7].

We find that the proposed control structure gives near optimal performance in the face of a realistic wind power distribution. As such, it incurs zero steady-state loss and inherently considers the trade-off between the HTO and high pressure and lye flowrate setpoints. It can therefore both extend the feasible operating range in the low load region and improve the efficiency in the case of high power availability. The proposed structure is further easier to implement than a Model-Predictive Controller and can be implemented in the fast-acting regulatory control layer, offering an important extension to current efforts towards flexible green Hydrogen production.

[1] J. Brauns, T. Turek, 2022, Experimental evaluation of dynamic operating concepts
for alkaline water electrolyzers powered by renewable energy, Electrochimica Acta

[2] J. Brauns, T. Turek, 2020, Alkaline Water Electrolysis Powered by Renewable Energy: A Review., Processes, Volume 8(2)

[3] P. Trinke, P. Haug, J. Brauns, B. Bensmann, R. Hanke-Rauschenbach, T. Turek, 2018, Hydrogen Crossover in PEM and Alkaline Water Electrolysis: Mechanisms, Direct Comparison and Mitigation Strategies

[4] R. Qi, X. Gao, J. Lin, Y. Song, J. Wang, Y. Qiu, M. Liu, 2021, Pressure control strategy to extend the loading range of an alkaline electrolysis system, International Journal of Hydrogen Energy, Volume 46(73)

[5] L. Cammann, J. Jäschke, 2023, PROCEEDINGS OF THE 33rd European Symposium on Computer Aided Process Engineering, (ESCAPE33)

[6] S. Skogestad, 2004, Control structure design for complete chemical plants, Computers & Chemical Engineering, Volume 28(1-2)

[7] I. Staffell, S. Pfenninger, 2016, Using bias-corrected reanalysis to simulate current and future wind power output, Energy, Volume 114(1)

[1] To be presented at the European Symposium on Computer-Aided Process Engineering, June 2023