Solar reflectance of most diffused flat surfaces is fairly constant for all daylight hours expect early in the morning and late in the afternoon. Hence, most standards rate and label the reflectance of roofing surfaces with a single value. However, some newly developed roofing materials, called directional reflective materials (DRMs), are designed to have dependent reflectance on incident sunlight angle. 
The Cool Roof Rating Council (CRRC) has investigated several approaches to label DRMs and have selected a single reflectance. However, the performance of the labelling approaches to estimate HVAC energy consumption of buildings has not been investigated yet.
In this study, an algorithm to simulate the mean hourly reflectance of DRM roof is developed using DOE-2.1E program. Afterwards, the HVAC energy consumption of building is simulated based on the calculated hourly solar reflectance. In addition, DOE-2 simulations were performed using the reflectance calculated by several labeling approaches, and their effects in predicting cooling, heating and total HVAC energy consumption of a building with DRM roof is assessed. 
The results indicated that the selected approach by CRRC (M2) is the most accurate model to represent the reflectance of DRMs. It has the most accuracy in estimating annual space cooling energy by an estimation difference of 1.7% (0.92 kWh/m2) in Houston, 1.1% (0.32 kWh/m2) in Sacramento and 1.8% (0.2 kWh/m2) in Montréal. M2 overestimates the space heating energy by overestimating reflectance of DRMs during winter. It tends to overestimate the annual space heating energy by as much as 1.8% (0.12 kWh/m2) in Houston, 0.5% (0.1 kWh/m2) in Sacramento and 0.3% (0.38 kWh/m2) in Montréal.