Multiparametric analysis of a single-stage submersible pump with emphasis on predictive maintenance using optical fiber sensing

Abstract

Electrical Submersible Pumps (ESPs) are widely employed in industrial applications that require high operational reliability and continuous operation under severe conditions. In this context, monitoring these machines becomes essential for the early detection of faults, reduction of unplanned downtime, and optimization of maintenance strategies. Therefore, this work proposes a multi-parametric monitoring strategy based on the integration of Fiber Bragg Grating (FBG) sensors and Raman-based Distributed Temperature Sensing (RDTS). The integration of these technologies enables the simultaneous acquisition of thermal and vibrational parameters of the ESP under different operating conditions. FBG sensors were employed for localized vibration measurements, allowing the identification of spectral components associated with the rotational dynamics of the machine and possible mechanical irregularities. In parallel, the RDTS system was used to obtain distributed temperature profiles along the optical fiber installed within the pump, enabling the spatial analysis of thermal gradients within the motor and regions subjected to higher heat dissipation. The proposed monitoring architecture employs independent optical channels for the simultaneous acquisition of signals from the FBG sensors and the RDTS system, enabling an integrated analysis of the thermal and mechanical phenomena occurring in the ESP. While the RDTS system provides the spatial distribution of temperature along the optical fiber, the FBG sensors enable the acquisition of dynamic signals associated with the vibrational behavior of the machine. To validate the proposed methodology, an experimental test bench was developed under controlled conditions, in which the pump was instrumented with the optical sensing system and subjected to tests under different flow rate conditions. The experimental results allowed the identification of spectral signatures associated with the rotational frequency of the motor–pump assembly, as well as components related to mechanical irregularities, such as rotor dynamic eccentricity. In the thermal analysis, the distributed temperature profiles enabled the observation of the spatial and temporal evolution of heating along the instrumented structure, allowing the correlation between the operating conditions of the pump and the thermal distribution of the equipment. Thus, the proposed monitoring strategy integrates thermal and vibrational information into a single optical sensing architecture, providing support for continuous monitoring and predictive maintenance applications in Electrical Submersible Pumps. Furthermore, the proposed methodology has potential applications in different industrial systems that require real-time multiparametric monitoring.