AML:FC

AML:FC - The single-cell model

The basic building block

The single-cell model is at the heart of the AML:FC. It is a building block for many other applications as well as an important device for model-based analysis of experimental data.

The heart of the fuel cell

The single-cell model comprises the anode-electrolyte-cathode assembly shown below. It is constructed from the basic elements of the AML:FC SOFC or AML:FC PEM libraries.

A more detailed example for an SOFC model can be seen on the technology page.

The single-cell model is constructed in the following way:

All key interactions in the anode-electrolyte-cathode assembly (or membrane-electrolyte assembly) are accounted for simultaneously, including detailed mass and heat transfer, reaction and electrochemistry effects.
Fluxes through the assembly are modelled in a 1-dimensional distribution.

Single cell modelscan be combined in x and y dimensions to create two or three-dimensional models when required.

A more detailed example for an SOFC model can be seen on the technology page.

The single-cell model constructed in this way:

takes into account all key interactions in the anode-electrolyte-cathode assembly (or membrane-electrolyte assembly) simultaneously, including detailed mass and heat transfer, reaction and electrochemistry effects.
models fluxes through the assembly in a 1-dimensional distribution. Many such single-cell models are combined in x and y dimensions to create two or three-dimensional models when required.

Validation: model-based data analysis and parameter estimation

A key use of the single-cell model is for model-based data analysis and parameter estimation from single-cell experimental data, in order to:

estimate parameters for key empirical values - for example deactivation kinetic constants
analyse the reliability of the experimental data based on parameter uncertainty:
- to determine whether further experimentation is required, and what data needs to be measured in order to reduce design risk.

By using single cell data it is possible to control conditions in order to isolate the effects being studied without introducing extraneous factors resulting from the interactions between multiple cells.

Model-based data analysis provides confidence information that can be used to determine whether additional experiments are require to refine the data. The model can also be used with model-based experiment design techniques to determine the optimal form of those experiments.

The following two plots are from an AML:FC single cell model that has been validated against experimental results. The plots show the simulated IVP curves against real experimental data before and after parameter estimation and model validation. As can be seen the fit for low current density has been improved by the parameter estimation.


Before parameter estimation	After parameter estimation

Deactivation

The single cell model is also used to study and quantify deactivation. In particular it is used to estimate deactivation parameters using data from accelerated deactivation experiments.

The basic form of the deactivation equation takes into account the history of production at a certain point in the MEA as follows:

Cell activity = 1 — local cumulative production * deactivation coefficient

Cumulative production is computed locally as a quantity distributed over the dimensions of the cell.

This takes into account the fact that different parts of the MEA deactivate at different rates, depending on the local conditions they experience - for example, temperature and composition history. The deactivation coefficient itself can be different for different locations.

Once deactivation coefficients have been determined using small-scale experiments, they can be applied to all cells in a stack or assembly and deactivation calculated over periods of months.

It is possible to use gPROMS's dynamic optimisation capabilities in order to determine the optimal values of design variables for minimim deactivation over the required lifecycle.

PSE can assist with deactivation studies (including the formulation of deactivation experimental procedure) through our ModelCare programme.