The favorable feed-in tariffs have rendered residential photovoltaic (PV) installations an attractive investment for homeowners. Due to the restricted space availability as well as the regulatory limitations for the allowable installed capacity, optimization of the modules' configuration becomes an important aspect toward maximization of PV investment efficiency. In this study, a design methodology for the optimal deployment of PV rows installed on flat-roofs is presented. The proposed methodology includes a photovoltaic energy production model that takes into consideration the PV self-shading losses and their effect on the energy production, while the optimization process is performed by considering the effect of rows geometry on the energy output and on the net present value. These optimization problems are formulated as constrained nonlinear programming problems and they are solved by using well-known deterministic methods. The results from a case study illustrate the importance of the interaction between the PV panels dimension and the geometry of the created PV rows with regard to optimal PV field designs under appropriate constraints. Further and as far as financial parameters like investment cost and feed-in-tariff are concerned, the results suggest that economic profitability is not always leading to energy maximization and thus suitable designs must be applied. (c) 2014 Elsevier B.V. All rights reserved.