Building energy simulation programs are powerful tools which are increasingly used at the design stage, for the assessment of compliance with energy efficiency regulations and for integration into energy management systems. A wider scope, a higher level of detail of building and thermal systems and the trend towards the use of dynamic simulations in real time are making these programs more computationally demanding. In this context, this paper presents a study of the computational time reduction in thermal simulations of buildings by applying equation tearing methods. These methods are employed to construct subsets of equations that can be independently solved with lower computational cost. The study was conducted at seven typical heating, ventilation, and air conditioning facilities. The computation time has been evaluated in five-minute intervals during a one-year period. The average saving ratio in processing time was 4.4; ratios ranged from 1.9 to 7.1. (C) 2016 Elsevier B.V. All rights reserved.