(115a) Dynamic Optimization of Proton Exchange Membrane Water Electrolzyers Considering Usage-Based Degradation

Electrolytically produced hydrogen (H₂) continues to gain traction as a means of producing low- carbon hydrogen. As compared to other low temperature electrolyzers, proton exchange membrane (PEM) systems are particularly promising because of their small size, ability to operate at higher current densities (≥ 1 A cm^−2), and ability to operate with a differential pressure between the anode and cathode, which enables production of high pressure H₂ product. Several techno-economic optimization studies have highlighted the importance of dynamic operation of PEM electrolyzers to minimize the levelized cost of hydrogen when utilizing electricity sourced from variable renewable energy (VRE) electricity sources that are co-located or supplied via connection to the electric grid. These studies reveal that cost-optimal dynamic operation involves operating operating at high current densities (≫ 2 A cm^−2) during time of abundant VRE supply or low electricity prices, as well as periods of idling or near-zero current densities (during high electricity prices or low VRE supply periods), that collectively results in smaller stack areas when compared to an equivalent steady state electrolyzer operating at current densities near 1-2 A cm^−2.

Though more efficient for capital and operating costs, dynamic operation tends to lead to faster stack degradation and shortened lifetime in comparison to electrolyzers operating at steady state [1]. Operation at high current densities leads to higher rates of stack degradation, while PEM elec- trolyzers left to operate at near open circuit conditions have shown accelerated degradation as well [1]. Most techno-economic studies overlook such use-dependent degradation that would inevitably result from dynamic stack operation [2]. Here, we develop a first-principles, 0-D electrochemical model for PEM electrolysis to simulate stack level operation over time while tracking key process variables, such as temperature fluctuations and species concentrations at the anode/cathode. To characterize use-dependent stack degradation in our modeling, we develop an empirical relation to quantify marginal degradation rates as function of current density that is able to approximate experimental trends observed in the literature related to accelerated degradation at low and high current densities [1, 3–8].

We embed the above-mentioned first-principle stack model and degradation correlation into an techno-economic optimization model to evaluate least-cost system design and operation under time- varying electricity prices. The cost-optimization model, a nonlinear program (NLP) implemented in Pyomo, considers the following important operating constraints: a) hourly production requirements for H₂, met through a combination of electrolyzer operation and on-site H₂ storage discharge that was previously charged from the electrolyzer output, b) operating temperature limits for the stack, and c) ensuring H₂ concentration in the anode is below 2 vol% for safe operation [9]. The model is evaluated for various scenarios of future electricity price conditions as well as electrolyzer capital costs and stack replacement costs, to understand the value of dynamic operation vs. steady-state operation considering stack degradation cost impacts.

The presentation will cover a summary of the analysis with a particular emphasis on characterizing the type and range of dynamic operation that would be most economically viable relative to static operation when considering degradation associated trade-offs and electricity price volatility.