Protection systems are usually implemented in mountains aiming to resist natural dangerous phenomena. As any other critical infrastructure, protection systems should always withstand and operate efficiently as they guarantee the safety of people and protect socio-economic assets. However, the efficacy of these systems decreases with the increase of the deterioration level of the interdependent components, which constitute them. To provide desirable operation over their lifetime, the management of protection systems is of paramount importance. A major key issue in such critical infrastructure management is to optimize the cost effectiveness of maintenance actions while maintaining a sufficient protection efficacy. This study proposes a decision-aiding model to assess different maintenance strategies applied to a protection system against debris flows. The model is constructed using physics-based stochastic Petri nets. It incorporates (1) a stochastic deterioration model, which is a surrogate model of a physics-based model developed for building deterioration trajectories of the system and (2) maintenance model that permits assessing the cost and efficiency of maintenance strategies. This study addresses the case of a debris retention system, in which the progressive filling of its basin by debris materials is modeled. This is followed by assessing several maintenance strategies concerning the cleaning of the basin. A simple sensitivity analysis is also carried out in order to check the effect of the uncertainty that invades the model’s inputs on maintenance decisions. A numerical analysis is performed using real data of the retention system located in the Claret torrent in France and subjected to debris flows over a period of 50 years.