Initial Design and Implementation of a Control System for a Hydrogen-Based Microgrid

Abstract

The increasing integration of renewable energy sources into modern power systems necessitates innovative control solutions, particularly in microgrids operating under off-grid conditions. This thesis presents the initial design and implementation of a control and monitoring system for a hydrogen-based microgrid, currently under development on the area of the Hamburg Observatory (Hamburger Sternwarte). The system serves as a prototype for a backup power solution at the TIGRE Observatory in La Luz, Mexico, an observatory frequently affected by power outages. To address the challenges of off-grid reliability, a hybrid energy storage architecture was developed, combining photovoltaic generation, lithium-ion batteries, electrolyzers, and a PEM fuel cell. The planned control infrastructure is centralized around a Raspberry Pi 4, which manages data acquisition via multiple communication protocols. Including CAN bus, Modbus RTU/TCP, and Ethernet, while MQTT serves as a higher-level, publish-subscribe communication layer. Time-critical sensor data and safety-relevant operations are delegated to Arduino-based microcontrollers. A custom, GPIO-based, interrupt-driven interface was implemented between the Raspberry Pi and the Arduino Due to enable a fast and error-minimized data transfer. The software framework was developed using a modular architecture and implemented in Python and C++, comprising over a dozen dedicated scripts. These include programs for interfacing with the key hardware components (e.g., fuel cell, electrolyzers, multimeters) as well as modules for data logging, plausibility validation, and real-time plotting. Structured classes handle data acquisition, normalization, and error detection. All scripts operate asynchronously via MQTT, enabling scalable and decoupled functionality. Core management routines coordinate parallel script execution, system diagnostics, and watchdog-supervised runtime monitoring. In addition, a modular EMS logic was implemented as an MQTT-subscribing component, forming the foundation for future automation of hydrogen production and consumption control. A comprehensive requirements analysis was conducted, addressing key aspects such as communication compatibility, fault management, system stability, and resource efficiency. The proposed energy management strategy prioritizes battery use and activates hydrogen-based systems for long-term load balancing. Overall, this work demonstrates the technical feasibility of a flexible, low-cost control system for hybrid microgrids and provides a foundation for future optimization, field deployment, and integration of additional components such as graphical interfaces and further implementation of sensor networks and interfaces.