Laboratory experiments on cross-shore beach morphodynamics are presented. A lightweight sediment (density rho(s) = 1.19 g cm(-3)) model is used in order to fulfill a Shields number and Rouse number scaling. This choice aims at correctly reproducing bed load transport as well as suspension dynamics. Terraces and barred beach profiles obtained in the experiments also present close similarities with profiles observed in the field. In order to question the concept of equilibrium beach profile, wave forcings conforming to a JONSWAP spectrum were imposed over long periods (up to more than a hundred hours). An average bottom evolution velocity is defined and used to determine when the profile reaches equilibrium. Usually, beach profiles are characterized according to the Wright and Short (1984) classification based on the Dean number W. This well-known classification is investigated and refined in the intermediate range, that is, for 1 <= Omega < 5. For W close to 1, a typical reflective profile is obtained. Terraces are obtained for the Omega = 2.5 cases. For Omega approximate to 3.7, the profiles exhibit two parts: a mild dissipative offshore slope producing low reflection and a steeper beach face with slightly higher reflection. The wave dissipation, velocity skewness, and acceleration skewness are computed from the free surface elevation time series. The dissipation and wave nonlinearities patterns are similar for similar equilibrium beach profiles, that is, with the same Dean number. Dissipation peaks coincide with bottom slope transitions as higher energy dissipation occurs with milder bottom slope sections. Besides, the uniformity of volumetric wave energy dissipation seems to concern only a limited zone of beaches with a widely developed surf zone.