Recent years have seen an increase in the number of proposed models to explain Earth's mantle dynamics: while two end-members, pure layered convection with the upper and lower mantle convecting separately from each other, and pure, whole mantle convection, appear not to satisfy all the observations, several addition models have been proposed. These models include and attempt to characterize least one reservoir that is enriched in radiogenic elements relative to the mid-ocean ridge basalt (MORB) source, as is required to account for most current estimates of the Earth's heat budget. This reservoir would also be responsible for the geochemical signature in some ocean island basalts (OIBs) like Hawaii, but must be rarely sampled at the surface. Our current knowledge of the mass- and heat-budget for the bulk silicate Earth from geochemical, cosmochemical and geodynamical observations and constraints enables us to quantify the radiogenic heat enrichment required to balance the heat budget. Without assuming any particular model for the structure of the reservoir, we first determine the inherent trade-off between heat production rate and mass of the reservoir. Using these constraints, we then investigate the dynamical inferences of the heat budget, assuming that the additional heat is produced within a deep layer above the core-mantle boundary. We carry out dynamical models of layered convection using four different fixed reservoir volumes, corresponding to deep layers of thicknesses 150, 500 1000 and 1600 km, respectively, and including both temperature-dependent viscosity and an instrinsic viscosity jump between upper and lower mantle. We then assess the viability of these cases against 5 criteria: stability of the deep layer through time, topography of the interface, effective density profile, intrinsic chemical density and the heat flux at the CMB.