Use of Predictive Analytics on a Critical Heated Pipeline

This paper presents a case study of the application of predictive analytics using a fiber-optic based temperature monitoring system on an electrically heat traced asphalt pipeline at a refinery in North America. A skin-effect trace heating system and a fiber optic Distributed Temperature Sensing (DTS) system was installed on an approximately 700m long section ("DTS zone") of an asphalt pipeline passing over various railroads and highways. The piping upstream and downstream of the DTS zone was only heat traced but not monitored by a DTS system (i.e., outside of the DTS zone). Based on the data collected by the DTS system, the system’s software analyzed the thermal behavior of the pipeline to identify weak spots in the thermal profile of the pipeline, predict potential product freeze at certain locations and the time to freeze in the absence of remedial actions. The system is also being utilized by the operators to analyze the product temperature entering and exiting the DTS zone to determine functionality and operating limits of the heat tracing systems outside the DTS zone.