Abstract: Predicting neutron behavior in advanced reactors using dispersed particle fuel, particularly configurations containing both fuel and burnable poison particles, is challenging due to pronounced double heterogeneity effects. This phenomenon, involving complex neutron self-shielding interactions between different particle types, significantly impacts local neutron flux and integral reactor parameters. Traditional homogenization methods lack accuracy for these systems, while high-fidelity Monte Carlo simulations are often too computationally expensive for routine design analyses. This study addresses this gap by developing, implementing, and validating an accurate and efficient deterministic neutron transport methodology within the Bamboo-Lattice code framework. The primary goal is to explicitly model the double heterogeneity effect in fuel elements containing both fuel (UO₂) and burnable poison (B₄C) particles. The methodology integrates advanced techniques. Resonance self-shielding was treated using a hybrid approach combining the online subgroup method for detailed energy dependence and the Sanchez-Pomraning method to capture spatial self-shielding within the multi-layered structure of individual particles. Subsequently, the method of characteristics (MOC) was employed to solve the neutron transport equation across the full energy spectrum. The MOC solver explicitly modeled the heterogeneous geometry using effective multi-group cross sections derived from the hybrid online subgroup and Sanchez-Pomraning approach, which incorporated detailed subgroup calculations for the resonance energy range. This integrated methodology was implemented in Bamboo-Lattice. Validation was performed against high-fidelity continuous-energy Monte Carlo (NECP-MCX) reference solutions for benchmark problems involving idealized fuel cells with dispersed UO₂ and B₄C particles in a zirconium matrix. The results demonstrate the methodology’s accuracy. The calculated effective self-shielded cross sections for key resonant isotopes agree with Monte Carlo results within 5%. The calculated effective multiplication factor (k_eff) shows excellent agreement, with discrepancies consistently below 300 pcm. Furthermore, detailed neutron flux distributions within the particles and matrix were accurately reproduced, confirming the method’s capability to resolve strong spatial heterogeneity effects. In conclusion, this work successfully presents a validated deterministic approach within Bamboo-Lattice for analyzing complex dispersed particle fuels. By accurately capturing double heterogeneity effects, it offers a reliable and computationally efficient alternative to Monte Carlo for reactor design and safety analysis, overcoming the limitations of traditional homogenization. This represents a significant advancement for the neutronic simulation of advanced reactor concepts.