Abstract
A resistor network model is developed for solid oxide fuel cell (SOFC) composite anodes, in which solid electrolyte grains, metal particles, and pores are considered on the same footing. The model is studied by a Monte Carlo simulation on a face‐centered cubic lattice, with a random distribution of the three components over the lattice sites. The concept of active bonds is used; the bond between a metal and an electrolyte site is conductive (reaction‐active) if the sites belong to clusters connected to the solid‐electrolyte membrane or metal current collector, respectively, and if the bond has at least one neighbor site which is a part of a pore cluster connected with the fuel supplying gas channels. Active bonds are characterized by an elementary reaction resistance, inactive bonds are blocking. The total inner resistance of the anode is calculated as a function of composition and the elementary reaction resistance, Rr, vs. ion transport resistance, Re (of a "bond" between two solid‐electrolyte grains). Compositions which provide the lowest inner resistance for a given Rr/Re ratio are revealed. Across‐the‐sample distribution of the current through the three‐phase boundary is investigated. The higher the Rr/Re ratio, the larger areas of the three‐phase boundary are used; however, if the ratio is low, the reaction occurs only very close to the anode /membrane interface to avoid ion transport limitations. A scaling law for the reaction penetration depth inside the anode, 
(where β ⩽ 0.5) is suggested in accordance with the Monte Carlo results. In line with the existing experimental data, the simulation and scaling estimates reveal the interplay between the reaction penetration depth and the anode thickness, which determines the thickness effect on the inner resistance.