(74a) Fiber Optic Distributed Sensing for Process Monitoring and Improvement

Distributed fiber optic sensing is a powerful monitoring technology that transforms optical fibers, most typically used for telecommunications, into sensing arrays with up to many thousands of independent measurement locations over distances ranging from a few hundred meters up to tens of kilometers. The spatial coverage and sampling density captured by these systems can offer unprecedented insight into process dynamics if utilized in an intelligent manner. The basic system architecture consists of a sensor box (“interrogator”) connected to an optical fiber that acts as a passive sensor array. This optical fiber can be incorporated into robust cable assemblies suitable for installation in harsh and high temperature environments common in the petrochemical industry. Simultaneous temperature (distributed temperature sensing – DTS), uniaxial strain (distributed strain sensing – DSS) and acoustic (distributed acoustic sensing – DAS) monitoring is possible on a single cable by connecting the appropriate interrogator, either at the time of system commissioning or by retrofitting for additional monitoring capabilities.

DTS, DSS and DAS are based on optical time domain reflectometry (OTDR) measurement techniques in which an incident pulse of laser light travels along the fiber. For each sampling interval of fiber, a small amount of light is scattered backwards and recaptured by the fiber waveguide in the return direction. Local variations of the backscatter waveform intensity and wavelength provide information on the state of the fiber at successive sampling intervals. The data is distributed in distance from the measured roundtrip transit time of the light. Through continuous analyses of the backscattered signal from successive incident pulses, dynamic profiles of temperature, strain and acoustics are realized as a continuous 2D function of recording time and distance along the fiber.

Data from distributed sensing systems can be analyzed in real-time using application specific software installed on edge computing platforms paired with the interrogator unit(s). These integrated sensor systems can provide automated monitoring and alarm generation with either analog (e.g. relay or 4-20 mA) or digital (e.g. Modbus TCP) integration with a plant DCS, allowing an operator to make informed control decisions rather than navigating large and complex datasets. Current systems can output threshold or rate of change alarms for temperature anomalies such as for early hot spot detection and mitigation or for informed power cable monitoring to optimize circuit loads without the risk of conductor overheating. Similarly, strain data or acoustic amplitudes of specific frequency bands can be analyzed in real-time for process optimization or asset integrity monitoring. By wrapping fiber around pipes, DAS can transform each wrapped section into a passive, non-intrusive, high-resolution flowmeter by monitoring the velocity of turbulent vortices.

Although distributed sensing has been successfully integrated into permanent process monitoring systems at several plants, technology adoption in the petrochemical industry is still in its infancy. Research and development efforts currently underway aim to expand applications from actionable data based on absolute values or thresholds, to more advanced indirect analyses based on artificial intelligence and machine learning techniques. It is expected that the rate of system deployments will significantly increase in the next few years as the suite of applications are further expanded.