(453b) Online Dynamic Observer of Stack Hydration Profile
AIChE Annual Meeting
2009
2009 Annual Meeting
Computing and Systems Technology Division
Dynamics and Control of Fuel Cells
Wednesday, November 11, 2009 - 3:35pm to 3:55pm
Stack water management during dynamic operation is key to meeting quality/performance, reliability, and durability (QRD) of fuel cell vehicles. Dry operation results in lower membrane hydration which increases protonic conductivity and impacts fuel consumption while wet operation results in electrode and channel flooding and impacts voltage instability and operation reliability. Moreover water management is a key dynamic parameter to achieve reliable freeze starts which is a focus of active development (both in academia and automotive industry).
The key to optimal control of membrane hydration is estimation of membrane hydration profile within the stack. Traditional approaches use sensors on gas streams to measure humidity of inlet air stream and online models to estimate membrane hydration based on process measurements (Temperature, pressure, air flow). Even though such models are mature in literature, their prediction is limited by accuracies and dynamic capabilities of sensors. Alternative approach is to directly measure stack's internal resistance as a high frequency resistance (HFR) measurement. Use of HFR to correlate membrane hydration is a popular and well understood technique in fuel cell and often used during design and material screening phase as lab-grade diagnostics.
However for online application stack HFR measurement is an average measurement for the stack and provides no information on the underlying hydration profile of the membrane. To avoid electrode flooding it is necessary to know the location of wettest region in stack and control to that. This work presents a model based approach to dynamically estimate stacks membrane hydration profile based on HFR and auxiliary sensors. Ability of dynamic observer to independently estimate changes in inlet RH or outlet RH will be illustrated. Also its ability to operate in wider window of stack operation will be demonstrated both in power and humidification domain. Finally its application within closed loop controls will be discussed.