One of the traditional approaches used to derive energy production estimates of wave energy converters relies on seastate description based on integral parameters, through the use of scatter diagrams to describe potential deployment sites. This study compares, using numerical modelling, energy production estimates based, on one hand, on the integral parameters Hm0 and TE to describe the wave resource and, on the other hand, on a fully spectral representation of the sea state. The wave energy converter considered is a generic version of Wavebob. The numerical model used is a linear potential flow theory boundary element method (BEM) code, running in time domain and which includes corrective terms to account for viscosity. The spectral description of the sea states is based on the hindcast database HOMERE. In one of the sea states considered, the difference in energy production estimates between the spectral approach and the approach based on integral parameters is 77%.