A multi-band microstrip antenna sensor for real-time, non-invasive detection of water-gasoline mixtures

Accurate detection of water contamination in gasoline is essential for maintaining fuel quality, operational safety, and equipment longevity in the oil and gas industry. Even small amounts of water can cause corrosion, combustion inefficiencies, and costly maintenance. To address this, a microstrip antenna-based sensor is proposed for real-time, non-invasive detection of gasoline volume fractions in water–gasoline mixtures. Designed and simulated using Advanced Design System (ADS) software, the sensor operates at three resonant frequencies—2.2, 4.5, and 5.1 GHz—enabling multi-band sensitivity. It is fabricated on a low-cost FR4 substrate with a compact footprint suitable for industrial deployment. Experimental validation was conducted using mixtures with water content ranging from 0% to 100% in 20% increments. Return loss (S₁₁) shifts corresponded directly to changes in dielectric properties, allowing accurate identification of mixture compositions. To improve prediction accuracy, Piecewise Cubic Hermite Interpolation (PCHIP) was used to expand the dataset, followed by nonlinear cubic regression. To improve prediction accuracy, PCHIP was applied to expand the original 6 experimental points into a denser dataset of 106 points, followed by nonlinear cubic regression. This approach achieved an RMSE of 0.047 on the full PCHIP-augmented dataset and 0.997 when evaluated solely on the original 6 experimental points in volume fraction estimation. Sensitivity analysis showed relative sensitivities of 14.98%, 19.25%, and 63.56% across the three frequency bands, with an average of 32.6%, outperforming many existing microwave resonator-based sensors. Compared to traditional laboratory-based or invasive sensing methods, the proposed sensor offers advantages in accuracy, miniaturization, cost-effectiveness, and ease of integration. Its single-layer planar design, fast response, and compatibility with standard PCB fabrication processes make it well-suited for real-time fluid composition monitoring in applications such as fuel quality control, pipeline inspection, and environmental safety.