Abstract Various mechanisms have been developed worldwide to support renewable energy sources. These mechanisms promote greater integration of renewable energy sources into power supply systems. Many support mechanisms can operate only in centralized power systems with well-structured market relations. As far as autonomous power systems are concerned, high expectations are placed on energy service contracting which offers a number of advantages such as instant investment, absence of risks for local authorities, fuel savings and efficient technical solutions. This paper presents a methodology for installed capacity optimization of autonomous photovoltaic systems under energy service contracting. The proposed methodology is based on the chronological method for calculating power supply systems using multi-year meteorological data sets. The study uses classical mathematical models of components of power supply systems: solar panels, solar inverters and direct current combiner boxes with allowance for operational restrictions. The internal optimization algorithm is described for the installed capacity and standard size of solar inverters and direct current combiner boxes. A hierarchical two-stage model for interaction between energy service company and regional authorities is proposed. The installed capacity of the Tyoply Klyuch photovoltaic system (Far East, Russia) has been optimized. The optimization process took into consideration nine types of solar panels: monocrystalline, polycrystalline and heterojunction panels. The reported value of the installed capacity is 1600 kW under the nine-year long contract. The annual fuel economy constitutes 300 tons, and a profit for the energy service company is US$ 0.428 million at the expiration of the contract. The methodology described in the present paper is universal and can be applied throughout the world.