Detection of Multiple Small Temperature Events Simultaneously on a Distributed Temperature Map

Measuring temperature is essential in many industrial applications not only to preserve equipment lifespan but also for security reasons. One solution that has grown exponentially in recent years due to its distributed measurement capability is the Raman-based distributed temperature sensors (RDTSs) that enable the temperature to be mapped at several points along tens of kilometers of fiber optic cables. In spite of being a technology already present in the industry, it shows no sign of abating with progress being achieved in basic interrogation, accuracy improvements, and novel applications. Currently, the scientific literature in the area of distributed temperature sensing exhibits a consistent body of knowledge based on the solutions that allow temperature measurements with high precision and resolution. However, these approaches fail in situations when the temperature event is less than the equipment's spatial resolution, so that it cannot be detected correctly. In this perspective, we propose and experimentally demonstrate in this study an image processing technique applied to the temperature profile generated by these sensors that will allow circumventing this bottleneck of RTDS technology. As a result, the simultaneous detection and correction of hot spots of sizes ranging from 15 m to 3 cm are achieved experimentally with 98.22% accuracy. These results obtained here place the RDTS technology at a new level of applicability, in which new industrial applications can be developed.