High resolution microscopy images are commonly used for studying anode layers and other components used in fuel cells. The performance of fuel cell materials is largely governed by their transport properties. Fluid, ions and electronic conductivity in particular depend on the microstructure. Understanding and modeling their morphology at the microscopic scale is therefore critical to develop new devices with improved properties. In this study, several methods are employed to model these media, based on three types of anode layers of different origins and aspect. A general methodology is used to represent materials made of three different phases, based on two random 3D sets which are independently chosen. The independent models are computed according to statistical measurement carried out on the experimental images. The best model, which reproduces the correlation function of the experimental images and other statistical features, is shown to model accurately the three types of anode layers investigated. The correlation function of the anode layers is also modeled, which can serves as the basis for generic models of these materials.