Fuel cell system design
Accelerating commercialisation with high-fidelity system modelling
Successful commercialisation of fuel cell technologies depends largely on how the fuel cell stack is integrated within the entire power plant system.
PSE's gFUELCELL advanced process modelling environment provides extensive system models within a dynamic simulation framework, allowing rapid system design and accurate analysis of performance under different operating conditions.
System models: exploring the decision space
Combining a high-fidelity, multi-scale fuel cell stack model with models of the entire power plant system provides an essential analysis tool for quantifying behaviour and performance at all scales.
Fuel cell powertrain undergoing a drive-cycle analysis
PSE's gFUELCELL environment provides unprecedented accuracy, from microscale phenomena to system-wide effects, to yield a wealth of accurate information about the system performance that cannot be obtained via experimentation alone.
In particular, rigorous models allow for tracking of gas and liquid state water in the membrane electrolyte assembly (MEA) layers and provide insights into the mechanisms for transport of water within the fuel cell. These are key factors affecting fuel cell performance and longevity.
The flowsheet above represents a typical automotive fuel cell system undergoing a drive cycle analysis. The dynamic model helps designers to understand relevant interactions such as the effect of recycle streams, and to design operating procedures for start-up, shut-down and handling of emergency situations.
First step: conceptual system design
In many cases it can be advantageous first to model a rigorous fuel cell within a simple system, such as the one shown above, in order to screen stack performance aspects such as transient water management and deactivation under various scenarios.
This allows designers to rapidly investigate the pros and cons of alternative system designs and control strategies.
System models: key attributes
Relevant attributes of gFUELCELL's system models are:
- Simulates in fully pressure-driven mode to ensure accurate representation of all key transient effects
- Includes computationally-intensive, complex models such as the PEM stack and humidifier to ensure full multi-scale accuracy
- Robustly simulates extremely dynamic operations such as drive cycles
- Accommodate hierarchical control structures using signal model libraries
The result is a high degree of predictive accuracy capable of reproducing pilot plant information and real operation.
Monitoring of all vital KPIs
Because the system model includes a detailed model of the stack, it is possible to monitor all vital stack key performance indicators (KPIs) during the driving cycle – including properties distributed in the through-plane and in-plane dimensions of the MEA.
The image to the right shows dynamic prediction of PEM water uptake during a dynamic system simulation (not in real time). This provides an unprecedented level of design and performance information that allows designers to rapidly explore the effects of design decisions, rank and screen stack and system design alternatives, and effectively design control systems to mitigate risk.