Latent storage-enhanced distributed temperature control in hydrogen microreactors

Microchannel catalytic plate reactors are a promising route for converting methane from geographically distributed sources (e.g., shale gas deposits) to hydrogen or liquid transportation fuels. Their capacity is easily scalable by increasing the number of units and thus well suited to distributed production needs. However, miniaturization inevitably reduces the number of available actuators and sensors, and the control of these inherently distributed systems presents challenges. In the present paper, we concentrate on autothermal microchannel reactors producing hydrogen via methane-steam reforming, and introduce a novel temperature control strategy based on the use of a layer of phase-change material (PCM) confined between the reactor plates. The PCM layer, which mitigates temperature excursions through melting-solidification occurring due to fluctuations in hydrogen production rate, acts as the distributed tier of a hierarchical control structure. The supervisory layer consists of a model-based feedforward controller. We also introduce a novel stochastic optimization method for selecting the PCM layer thickness (i.e., for distributed controller “tuning”). The proposed approach is tested in simulations carried out on a detailed 2D reactor model, showing excellent disturbance rejection performance.