This paper investigates the possibilities of noise prediction in Heating Ventilation and Air Conditioning (HVAC) systems using semi empirical scaling laws. An approach is presented where the general noise reference spectra are combined with Reynolds Average Navier Stokes (RANS) simulations. Focus is at applying the suggested noise prediction approach to common HVAC components but also to discuss the properties of the prediction model, e.g. radiation characteristics and chosen reference spectra. A model is presented, using a momentum flux assumption of the noise sources, which is validated by a range of HVAC components of both high and low pressure loss.