Process Systems Enterprise Limited
email this page print this page
pdf overview

Advanced Model Library for Fuel Cells (AML:FC)

Advantages of the AML:FC

Typical fuel cell modelling applications

AML:FC at-a-glance …

  • highest fidelity available
  • steady-state and dynamic
  • complete flexibility of configuration
  • detailed representation of all key phenomena
  • full model validation against laboratory data
  • combined gPROMS and CFD
  • speed and robustness
  • flowsheeting for systems design
  • control design in gPROMS or Simulink®
  • design of experiments
  • dynamic optimisation
  • access to all standard gPROMS facilities

There are many general benefits arising from the systematic use of advanced process modelling in fuel cell component and system design. This page lists the some typical applications and specific advantages of the AML:FC over other fuel cell models on the market.

AML:FC modellers can take full advantage of all the standard facilities of gPROMS, such as the powerful built-in parameter estimation, optimisation and experiment design techniques and links to CFD packages for hydrodynamic modelling.

Some examples where this provides significant advantage are:

  • estimation of key system parameters from single-cell experiments, and related model-based data analysis for risk assessment
  • deactivation analysis and management
  • detailed design of flow channels taking into account all relevant fluid flow and mass transfer phenomena
  • determining optimal PEM water management strategy
  • detailed mechanical design of cell stacks
  • integration of fuel preparation systems, including heat integration and start-up dynamics
  • control system design.
There are many more typical applications. Once a high-fidelity model of cell, stack or system is available it can be used to investigate and optimise virtually any aspect of design and operation, with physical testing used only to verify final results.

Specific advantages of the AML:FC over other fuel cell models:

Some key advantages are listed here. For more concise information see the comparison with other tools.

Highest-fidelity models

The AML:FC models are constructed to the highest standard using state-of-the-art modelling techniques such as Maxwell-Stefan multicomponent diffusion techniques to take into account all relevant interactions between components. Because of the way that gPROMS works, all of these effects are considered simultaneously during solution.

For enhanced accuracy distributed systems modelling is used throughout, and users can increase or decrease the level of discretisation as required.

Steady-state and dynamic

Many modelling tools cater primarily or exclusively for steady-state modelling. However most critical design decisions require analysis of performance during transient operation.

gPROMS makes no distinction between steady-state and dynamic models - it is designed to handle both with equal ease. The AML:FC models are all constructed for both steady-state and dynamic operation.

Using gPROMS's powerful dynamic modelling and optimisation features it is possible to accurately model essential time-varying interactions such as performance during load change or optimise operating procedures, for example to determine optimal start-up trajectories.

Complete flexibility of configuration

The AML:FC includes a range of different component models enabling construction of virtually any configuration of cell models. It is for example possible to create 1-D, 2-D, 3-D representations; different flow channel architectures; co- or counter-current flow and to implement these within a stack of tens or hundreds of cells. Further is is possible to incorporate fuel preparation systems, ancillary equipment and control system.

Detailed representation of all key phenomena

It is possible to solve highy complex models in gPROMS robustly and quickly, no detail has been spared.
  • electrolyte swelling
  • platinum distribution
  • agglomeration effects
  • water condensation and diffusion in all membrane electrode assembly layers
  • "water drag" effect on ionic transport

Model validation against laboratory data

Although good models are vital for accurate process simulation these models have to be validated against experimental data in order to adjust model parameters to reflect reality.

gPROMS's state-of-the-art model validation tools allow estimation of multiple model parameters from steady-state and dynamic experimental data, and provide rigorous model-based data analysis.

Combine the best of Advanced Process Modelling and CFD

The AML:FC-FLUENT Hybrid Modelling Interface for detailed cell stack design allows the user to link a gPROMS model of the anode-electrolyte-cathode assembly to a FLUENTŪ model of the flow channels.

For complex flow channel geometries (for example serpentine flow paths) an AML:FC membrane model, validated against experimental data, can be connected to a detailed hydrodynamic model of the channels, accounting for their exact geometry.

The combined co-simulation or hybrid simulation model provides the ultimate accuracy for mechanical design considering all interactions simultaneously.

Speed and robustness

The AML:FC makes use of gPROMS solvers, renowned for their speed and robustness. This enables quick and reliable solution of complex cell and stack models.

The AML:FC-FLUENT Hybrid Modelling Interface incorporates many advanced model reduction techniques that can reduce the time required for complex stack modelling by orders of magnitude. Consequently, accurate modelling of heat and concentration effects in stacks involving tens or even hundreds of individual cells is easily achievable.

Flowsheeting for systems design

An essential part of fuel cell development and design , particularly for systems that include integrated reforming, involves studying the complex interactions between different system components.

The gPROMS ModelBuilder® environment provides all necessary facilities for system modelling. ModelBuilder allows you to easity create and maintain an entire system flowsheet by combining models from different libraries, including control models.

This allows the user to study and optimise system interactions and analyse the dynamic behaviour of the entire system, as well to perform control design or design an optimal operating policy.

High-fidelity non-linear models for control design in MATLAB and Simulink

The high-fidelity process models constructed using the AML:FC can be exported to execute within The MathWorks Inc.'s MATLAB® and Simulink® control design environments. This is done using the gPROMS Objects for MATLAB and Simulink, gO:MATLAB and gO:Simulink respectively.

Control designers can use the same models as used for the process design, speeding up design and verification of the control scheme and resulting in a better control design.

Use of gPROMS to design optimal experiments

TheAML:FC models can be used with gPROMS's model-based experiment design capabilities to design experiments that generate the maximum amount of parameter information from the minimum number of experiments.

This helps to accelerate development, minimise cost and reduce design risk, as well as to integrate R&D experimentation activities with engineering design.

Optimise directly - not by trial-and-error

gPROMS's comprehensive tools for the optimisation of design variables and operating policy enable direct solution of design challenges instead of lengthy trial-and-error simulation.

Using dynamic optimisation it is, for example, possible to design:

  • optimal start-up policy taking into account system and material constraints
  • determine maximum rate-change constraints for load transitions
  • determine optimal equipment sizes for desired performance
directly rather than by repeated simulation.