Multispectral Sensing for Detection and Volume Prediction of Dielectric Fluids

This project addresses the critical challenge of detecting and quantifying spills of dielectric fluids in marine environments. These fluids are essential insulators and coolants for high-voltage offshore equipment, such as power transformers. As offshore wind energy and electrical infrastructure expand, the risk of accidental spills increases. However, common dielectric fluids (like Hyvolt, Midel, and FR3) are transparent and possess unique properties that make them difficult to detect and measure with traditional optical sensors, complicating spill response efforts.

Previous research (OSRR Project 1163) confirmed that thermal and near-infrared sensors could detect these fluids, but the studies were limited to qualitative assessments of dyed, recycled fluids. A significant knowledge gap remains in the ability to quantitatively measure and map spills of undyed, transparent fluids under real-world conditions.

The primary objective of this 24-month, $353,422 project is to develop, validate, and operationalize a multispectral remote sensing system that can accurately detect, characterize, and predict the volume of transparent dielectric fluids on the ocean surface. The project will advance from two-dimensional qualitative mapping to three-dimensional, quantitative volume prediction.