For most applications thermal requirements are relatively constant whereas solar insolation is a highly variable energy source. Most water distillation plants, for example, cannot operate with variable heat input because the levels, pressure drops, and thermal gradients in the system must be kept constant in order not to degrade the overall system operating efficiency. The generation of electricity, both for base load and peak load, must follow a.time variation coherent with the power usage, not necessarily similar to the insolation curves. Space heating and cooling must also be able to operate independently from the variations of solar input. Generally, the load curves for the above applications have winter peaks which are exactly opposite to the peaking of the solar input. Insolation fluctuations belong to two categories: a determinist ic componen t which can be calculated on the basis of purely astronomic grounds; and a random fluctuation due to the stochastic properties of the atmospheric turbidity. The superposition of the thermal load requirement for a particular application on the insolation input profile for the site almost inevitably produces a substantial phase lag. To make up this misfit of load and supply, the most obvious solution involves the use of fossil fuel (oil, coal, gas etc.) or nuclear heat to fill the gaps, using the solar heat input as a means of saving a fraction of the fuel cost. For most applications, this fraction ranges from 30 70%. This, of course, entails a doubling of the heat input of the power application, with t h e solar part staying inoperative for long stretches of time, during which depreciation costs are incurred. The ideal system would be that capable of collecting solar energy in spring and summer and releasing it in the winter. Long term storage of solar heat in the ground approaches this ideal. The objective of this s tudy was to develop a procedure for the choice, simulation and evaluation of seasonal storage of solar energy collected by means of fiat-plate collectors. The main steps of this s tudy were: a. development of a stochastic model of global incident radiation based on the statistical analysis of past recorded hourly global insolation values; b. analysis of thermal transients in the soil exposed to a cyclic thermal input; c. development of a physical and numerical model for long-term heat storage; d. development of a comprehensive code to simulate and select the components in a complete space-heating system including collectors, storage, heat pump, and radiators.