The prediction of food freezing time has been one of the most relevant parameters for the design of freezing systems and the estimation of frozen food quality. However, the growing demand of frozen food and new energy efficiency standards in recent years, have generated interest in methodologies that describe precisely the evolution and distribution of temperatures and energy inefficiencies, such as conjugate food air heat transfer models and exergy destruction analysis. In this work, the freezing time of meat inside a domestic freezer is determined using a conjugate three-dimensional (3D) model with the turbulent k-epsilon model for the airflow and the apparent enthalpy methodology for the phase change of water in the food. Additionally, a local exergy destruction analysis is performed to quantify the irreversibilities produced by viscous dissipation and heat transfer during the freezing process. The results obtained shows that the turbulent model describes properly the airflow velocity, obtaining a non-dimensional heat transfer (Nu) 64% higher than laminar model. Also, the 3D model describes precisely the physical domain, obtaining a Nu 28% higher than the 2D model. Consequently, the turbulent 3D model obtains the best correlation with experimental results from literature. The exergy analysis determined a total exergy destroyed of 187 (W), mainly during the first hour of the freezing process. Two considerations were proposed to reduce the internal inefficiencies: controlling the airflow efficiently with baffles around the food and increasing the heat transfer rate by faster mechanisms, such as radiative or forced convective heat transfer. (C) 2021 Elsevier Ltd and IIR. All rights reserved.