This paper describes a general methodology for modelling renewable energy plants, in both discrete and continuous time domains, based on field measurements, for control design purposes. The plant model is obtained by applying standard identification techniques to the measured input-output data. It aims to accurately capture the delays and dynamics of the plant in response to sinusoidal power set-points within the frequency range relevant to small-signal electromechanical oscillations (0.1–1.5 Hz). This paper focuses on the communication delays between the central controller and the generation units, as well as the computation time of the devices involved in the centralised plant control. The proposed methodology is applied to an operational medium-scale photovoltaic plant connected to the distribution system of Spain. The identified models are validated against the real system using time-domain simulations and Bode plots, showing an adequate performance in the frequencies of interest.