In solar power tower (SPT) systems, selecting the optimum location of thousands of heliostats and the most profitable tower height and receiver size remains a challenge. Given the complexity of the problem, breaking the optimisation process down into two consecutive steps is suggested here; first, a primary, or energy, optimisation, which is practically independent of the cost models, and then a main, or economic, optimisation. The primary optimisation seeks a heliostat layout supplying the maximum annual incident energy for all the explored combinations of receiver sizes and tower heights. The annual electric output is then calculated as the combination of the incident energy and the simplified (annual averaged) receiver thermal losses and power efficiencies. Finally, the figure of merit of the main optimisation is the levelised cost of electric energy (LCOE) where the capital cost models used for the LCOE calculation are reported by the System Advisor Model (SAM)-NREL and Sandia. This structured optimisation, splitting energy procedures from economic ones, enables the organisation of a rather complex process, and it is not limited to any particular power tower code. Moreover, as the heliostat field layout is already fully optimised before the economic optimisation, the profiles of the LCOE versus the receiver radius for the tower heights explored here are sharp enough to establish optima easily. As an example of the new procedure, we present a full thermo-economic optimisation for the design of the collector field of an actual SPT system (Gemasolar, 20 MWe, radially staggered surrounding field with 2650 heliostats, 15 h of storage). The optimum design found for Gemasolar is reasonably consistent with the scarce open data. Finally, optimum designs are strongly dependent on the receiver cost, the electricity tariff and the assumed maximum receiver surface temperature. (C) 2016 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license.