This work studies in detail the photo-Fenton degradation process of Paracetamol (PCT) on an annular pilot-plant reactor using Computational Fluid Dynamics (CFD). A cylindrical lamp emission model was originally implemented over the structure of the OpenFOAM(R) platform and a multicomponent reaction mixture model was used to compute the temporal evolution of the different species at each point of the reactor. Once the proposed model was experimentally validated, the influence of different operating conditions (i.e. different strategies for hydrogen peroxide (HO) dosage, use of low recirculation flow rates (Qr), and a completely uncovered lamp setup) was studied. The results of the analysis showed that a double addition of HO (50% before the tank and 50% before the reactor) significantly reduces the reaction times of the process. Moreover, the overall PCT degradation rate does not change when Qr is increased, thus allowing the system to be operated with a recirculation flow three times lower than that the one used in the experiments. Thereby, the developed model allows identifying the reaction conditions that maximize the overall PCT conversion, making efficient use of HO (main chemical reagent) and reducing the electrical energy consumption (recirculation flow) by operating the system under conditions present in large-scale photochemical reactors.