A comparative study of passive solar building simulation using HOT2000, TRNSYS14, NETSPEC

Thermal behavior of several conventional and passive solar buildings is analyzed using three quite different simulation packages, HOT2000, TRNSYS14, NETSPEC. HOT2000 was developed for the Canadian R-2000 home program. It uses a steady-state model with mean monthly weather data. TRNSYS is a detailed transient simulation program from the University of Wisconsin which solves the heat transfer differential equations numerically with short time steps of 1 hour or less. NETSPEC is a simplified network program from Trent University, with time variation represented by exponentially-decaying modes. For each program, the amount of auxiliary heat required to maintain the living area at minimum 21{dollar}spcirc{dollar}C is computed for 14 different versions of four building models, for four winter months in Ottawa (December to March).; Generally the three programs give similar results. HOT2000 consistently predicts the highest auxiliary heats, especially in March. TRNSYS gives the lowest heating load in all unvented models with small windows. NETSPEC gives the lowest auxiliary heat in most vented models and unvented models with large windows in February and March. In all cases large windows are net energy losers according to HOT2000, and net winners according to NETSPEC. TRNSYS is the most detailed program and probably the most accurate. It shows windows as winners or losers in different cases. Thus HOT2000 probably underestimates useful solar gain, while NETSPEC overestimates it. In most cases, TRNSYS is nearer to NETSPEC than HOT2000.