This paper presents a discussion about an alternative calculation scheme that can be used to model nuclear fuel assemblies under water. In criticality Monte-Carlo multigroup calculations performed by French industrials, heterogeneous media (such as fuel assemblies or more generally any type of arrays of fissile material in water) are mostly modelled as a homogeneous medium whose nuclear cross sections are adjusted to match those of the original heterogeneous system. These cross sections are obtained using a preliminary flux calculation based on deterministic methods. For a fuel assembly, the simplest deterministic method is to consider a simple cell (fuel oxide cylinder surrounded by clad and water) with a moderation ratio determined by considering that the water around fuel rods and inside empty slots is homogeneously distributed all over the assembly section. This simplified calculation scheme, very useful for parametric studies and for modelling missing fuel rods which position is not known, allows fast multigroup calculations and only requires knowing the number of fuel rods in the section. However, in some cases, this simplified approach can lead to underestimate the assembly reactivity. Indeed, the moderation ratio is generally heterogeneous within the assembly section, particularly for boiling water reactor types assemblies for which empty slots are unequally dispatched within their section. Moreover, the modelling of the water located outside of the assembly section can lead to mis-estimate the moderation of the external fuel rods. The purpose of this paper is to present an alternative simple approach to use preliminary deterministic calculations to generate homogeneous cross sections for multigroup Monte Carlo codes, in order to avoid two dimensional deterministic calculations which limit the parametric studies possibilities and are computer-time consuming. This paper will firstly remind generalities about the problematics regarding missing fuel rods in fuel assembly modelling. Then, results obtained with the alternative calculation scheme will be presented, with an explanation of the discrepancies compared to other calculation schemes. At last, a discussion about the bounding trait of this alternative scheme and the parameters that can have an influence on its behaviour will be presented.