Abstract
A method for characterizing the response of a PEMFC to transients in the CO concentration in the fuel is presented. The method uses the wellβknown First Order Plus Dead Time (FOPDT) model to separate the system effects from the poisoning characteristics of CO as a function of concentration and dosage. The method is more representative of the data and should be used as an alternative to a simple slope measurement for reporting the experimental data of the poisoning and recovery processes. The results characterize a specific MEA by showing that the time constant for the poisoning process is extended from 4 β 48 s when 15% air was injected with an anode stream of 3,000 ppm CO/H~2~. For the recovery process, a 15% airβbleed reduces the dead time from 122 β 36 s and the time constant from 89 β 18 s. In addition, the gain is shown to be relatively independent of the high CO concentrations but it is affected by air bleeding. The gain changes from β0.10 to β0.08 mV/ppm CO for a poisoning process and from 0.09 β 0.08 mV/ppm CO for a recovery process with 15% air bleeding in the 3,000 ppm CO/H~2~. In general, the dead times and time constants for the poisoning process decreased as the CO concentrations increased at equal CO dosage. On the other hand, these parameters (dead time and time constant) for the recovery process increased as the CO concentration increased at equal CO dosage. The method is critical for comparison of the dynamic response of MEAs between laboratories and for the design of reformate fuelβprocessing systems.