Fuel cell system optimization
Team organizer: Rochdi Trigui, CR HDR
Different levels of control and optimization
In transport or stationary applications, hybrid systems including Fuel Cells have a level of complexity such that a system approach involving modeling is required. In these systems, several control levels are to be designed and optimized. As shown in Figure 1, for a vehicle application of a FCS, a level of local control is required for each subsystem to ensure proper operation. This controller communicates with an overall supervisor who is responsible for generating at each moment appropriate instructions for different subsystems.
Optimization criterion or criteria of the entire system (transport or stationary) can occur at several levels:
- Optimization of the design (or sizing)
- Optimization of local control
- Optimization of the overall control
The optimization criterion can be as simple as minimizing the consumption of hydrogen for a given use, but can include other aspects such as minimizing transitional demands of the most vulnerable components to improve lifetime. In the optimization design, the use of resources from manufacturing to recycling the entire system should ideally be taken into account.
Optimization of PAC
For the particular case of a subsystem of a FCS used alone or in a global hybrid system, the optimization actions are similar to what is mentioned above, meaning the optimization of subsystem design for given load specifications (cell core, hydrogen supply, air supply, cooling, etc..) and the optimization of their control. The system shall allow the battery to function in the best possible conditions: at a suitable temperature (for PEMFC: avoiding water condensation or drying of the electrolyte membrane, etc. for SOFC: maintaining the power generation, avoiding the anodic reoxidation, etc.), at a constant pressure (for PEMFC: with slight pressure differences between the anode and the cathode, for SOFC limiting load losses that potentiate the loss of sealing) for optimal performance, etc..
The control must also ensure that the battery and its subsystems will not reach operating points that may lead to a failure (see Area 3). FC control must necessarily go through an understanding of the whole system and the subsystem components (see Areas 1, 2 and 3). The objectives of this area are therefore relying on the results of areas 1, 2 and 3, to develop innovative control strategies to ensure both performance requirements (dynamic, performance, durability, …) and operating constraints (range of temperatures, humidification, dynamic current rise …).
Optimization of a hybrid system including a PAC
For a given use, energy and environmental performance of a system including a hybrid or fuel cell depend strongly coupled three dimensions:
- The topology considered (storage elements, number of converters, etc.);
- The sizing of the components;
- The (local and global) command.
On the one hand, for a given topology of a system and sizing of its components, optimized global commands based on one or more criteria can be developed. On the other hand, if the topology and type of command are fixed, we can try to optimize the size of components for a given use. Ideally, a global optimization including the three dimensions as mentioned above would be the most relevant.
Given the expected complexity, the work in this research area is a result of two timing schedules:
– In short term, develop optimized control for multi-consumer and multi-source systems including fuel cells by generalizing the work for hybrid systems with two sources. Methods such as dynamic programming, the principle of Pontryaguin minimizing or online optimization based on fuzzy logic can be used, developed and optimized.
– In medium and long term, develop a global optimization approach merging both design and control aspects. This optimization is necessarily based on multi criteria to reduce the environmental impact of the studied solutions. In the case of a detailed understanding of the use of the system, such a process may not be inspired by conventional design approaches, which are generally based on the peak power, leading to a design suitable for use.
This optimization approach must significantly rely on accurate modeling of the subsystems in the overall hybrid system. Modeling of fuel cells and its ancillaries must be a favored theme of exchange with other areas of the federation. For the modeling of other components, existing collaborations with laboratories at national and international level are to expand and strengthen.