Investigation and Modelling of DC Microgrid for Electric Aircraft

Abstract

To further promote the process of decarbonization, electric aircraft have emerged as a promising area of research in aviation electrification. Among various approaches, fuel cell-powered all-electric aircraft stand out as a key direction, requiring in-depth investigation and development of critical technologies. These include system configuration design, cryogenic power conversion systems, superconducting machines with specialized control strategies, fault protection mechanisms, and energy storage systems that integrate supercapacitors and batteries. In this context, this thesis focuses on the comprehensive study and advancement of these pivotal technologies to support the development of electric aircraft.
Fuel cells serve as the primary energy source in this system, supplying thrust power to the constant power load. To ensure stable operation, energy storage systems play a vital role in maintaining system power balance and regulating the DC link voltage. In this thesis, the multiple source DC network under droop control in EA is modelled and simplified. FC stacks, as the primary power source is equivalent as a controller current source to track the load change. The slow response rate characteristic of the FC is considered in the analysis. Based on this model, a stability boundary is derived using the Jacobian matrix and eigenvalues of the system. Sensitivity analysis of key parameters, such as droop settings, DC capacitance, and line inductance, is conducted based on bifurcation theory. A flowchart for designing a stable DC MG for EA is provided. Key specifications, such as DC capacitance, line inductance, droop settings and load profile during the flight should be considered at the early design stage for industrial implementation.
By modeling the multiple-source DC network with fuel cells as current-controlled sources and incorporating their inherently slow response, this research addresses critical real-world dynamics that impact system stability under load variations. The derivation of stability boundaries using Jacobian matrices and eigenvalue analysis equips engineers with quantitative tools to predict and avoid instability. Sensitivity analyses based on bifurcation theory further inform how key parameters—such as droop control settings, DC link capacitance, and line inductance—affect system stability, enabling optimized component selection and control strategy tuning early in the design phase. The provided design flowchart serves as a practical framework to guide industrial implementation, ensuring that vital specifications and operational conditions, including load profiles throughout flight, are systematically considered. Overall, these contributions facilitate the development of robust, fault-tolerant, and efficient DC microgrids for electric aircraft, reducing development risks and accelerating certification processes while improving operational resilience in complex aerospace environments.
An essential component of the propulsion system is the cryogenic DC/DC converter, which ensures stable DC bus operation for fuel cells and batteries. A cryogenic bidirectional DC/DC converter prototype, employing SoLid Cover IGBTs, has been designed and built as part of this study. Half bridge IGBT modules (A0-VS122PA200M7-L756F70) based on SoLid Cover (SLC) technology are capable of operating at temperatures as low as 110 K. The nanocrystalline core SC2049M1 has been tested under conditions where it is fully immersed in liquid nitrogen for cooling. The performance of this cryogenic converter in both buck and boost modes is rigorously analyzed and experimentally validated. Running the designed DC/DC converter prototype in cryogenic temperatures can effectively reduce losses and improve efficiency. When the DC/DC converter prototype operated at cryogenic temperature in buck mode tests, the power loss was reduced by nearly 40% and the efficiency was improved by 1.5% at 10 kW compared to operation at room temperature. During boost mode tests, the efficiency of the cryogenic DC/DC converter prototype was improved by 1.8% at 10 kW when cryogenic cooling both the IGBTs and the inductor.
The significant reduction in power loss— up to 1.5% in buck mode and 1.8% in boost mode translate directly to improved energy utilization, extended component lifetimes, and reduced thermal management requirements. Furthermore, the validation of cryogenic-compatible magnetic components, such as the nanocrystalline core inductor, supports the integration of high-performance passive elements optimized for low-temperature operation. Collectively, these advancements pave the way for more compact, lightweight, and efficient power electronics systems essential for next-generation electric propulsion architectures, ultimately accelerating industry adoption of cryogenic and superconductive technologies in aviation.
Ensuring the safety and reliability of electric propulsion systems is another critical aspect of electric aircraft design. This thesis conducts a system-level analysis and characterization of potential electrical faults within electric aircraft DC networks to inform fault management strategies. Equivalent circuit models are developed to represent various fault transients. For P-P faults, time-domain curves of fault currents are derived and characterized for different fault locations, with the peak magnitudes and their corresponding occurrence time clearly identified. For P-G faults. the neutral point-to-ground voltage serves as a reliable fault indicator for distinguishing different grounding fault scenarios. Additionally, the selection of HRG limits the magnitude of the grounding fault current. Recommended HRG settings are provided to corporate with the undervoltage relay or overcurrent relay. The analytical results verified through simulation models based on OPAL-RT real-time simulators. Controller-in-the-loop experiments are conducted to validate the proposed fault management strategies at the system level. To minimize the time delay between fault detection and fault isolation, a higher sampling frequency is required for the current and voltage sensors. Additionally, a higher control frequency is needed for the controller unit; however, it must not exceed half of the sampling frequency.
The findings presented have significant practical value for the electric aircraft industry, particularly in enhancing the safety, reliability, and efficiency of on-board electric propulsion systems. First, the development of accurate equivalent circuit models for fault transients enables engineers to predict fault behaviour precisely, facilitating timely and effective fault detection and isolation, which is critical to preventing cascading failures or system damage. The identification of peak fault current magnitudes and their occurrence times aids in the design of protective devices. Using neutral point-to-ground voltage as a reliable indicator for grounding faults, combined with optimized HRG settings, allows for better fault discrimination and current limitation, thereby reducing the risk of excessive fault currents damaging sensitive aircraft components. The integration of these analytical insights with real-time simulation platforms like OPAL-RT supports rapid prototyping and validation of fault management strategies, accelerating development cycles while ensuring system robustness. Finally, the recommendations on sensor sampling and controller frequencies provide critical guidance for the design of measurement and control systems that balance responsiveness and stability, minimizing detection-to-isolation delays and enhancing overall system resilience.
Together, these results provide a comprehensive foundation for designing safer, more reliable electric aircraft power systems that meet stringent industry standards and operational demands.

Date of Award	10 Sept 2025
Original language	English
Awarding Institution	University of Bath
Supervisor	Vincent Zeng (Supervisor) & Xiaoze Pei (Supervisor)