Moisture analysis of ethanol- hydrocarbon blends by Vis-NIR spectroscopy

Importance of the Topic

Water contamination in ethanol-hydrocarbon blends leads to corrosion, engine wear, and reduced component lifetime in modern fuel systems. Accurate moisture control is critical as even trace free water can cause operational issues. Vis-NIR spectroscopy offers a rapid, non-destructive alternative to traditional Karl Fischer titration, enabling faster raw material inspection, process monitoring, and final product quality control.

Objectives and Study Overview

This study evaluates the capability of visible near-infrared (Vis-NIR) spectroscopy to quantify water content in ethanol-hydrocarbon blends. The performance of Vis-NIR is compared against established coulometric Karl Fischer titration standards, aiming to develop a robust predictive model suitable for routine analysis.

Methodology and Instrumentation Used

A total of 51 ethanol-hydrocarbon blend samples covering moisture levels from 0.0026% to 1.2016% (determined by KF titration) were analyzed. Spectral data were acquired in transmission mode across 400–2500 nm using a Metrohm NIRS XDS RapidLiquid Analyzer with 2 mm quartz glass cuvettes. 47 samples formed the calibration set and 4 served as an external validation set. Vision Air 2.0 software managed data acquisition, processing, and model development.

Instrumentation

• NIRS XDS RapidLiquid Analyzer (Metrohm code 2.921.1410)
• Quartz glass cuvettes, 2 mm path length (Metrohm code 6.7401.210)
• Vision Air 2.0 Complete software (Metrohm code 6.6072.208)

Key Results and Discussion

A Partial Least Squares regression using six latent factors and focusing on water absorption bands at 1380–1670 nm and 1875–1990 nm yielded excellent agreement with reference values (R² = 0.999). Calibration and validation metrics were as follows: SEC = 0.0097%, SECV = 0.0116%, and SEP = 0.0120%. External validation samples demonstrated prediction errors within ±0.0120%, confirming model robustness.

Benefits and Practical Applications

• Significantly faster analysis compared to coulometric Karl Fischer titration
• Elimination of toxic solvents and reduced operator training
• Suitable for inline or at-line monitoring in fuel production and quality control
• Cost-effective solution for routine moisture analysis

Future Trends and Potential Applications

Advancements in chemometric algorithms and instrument miniaturization may expand the applicability of Vis-NIR spectroscopy. Integration with process analytical technology frameworks and development of portable analyzers could enable real-time moisture monitoring. Machine learning approaches may further enhance predictive power and adapt models to varying blend compositions.

Conclusion

Vis-NIR spectroscopy provides a reliable, rapid, and non-destructive method for quantifying water content in ethanol-hydrocarbon blends. The developed PLS model achieves high accuracy and precision, positioning Vis-NIR as a practical alternative to traditional KF titration for fuel moisture analysis.

References

• [1] MyCleanDiesel.com. Problem: Water in Diesel Fuel. Available at http://www.mycleandiesel.com/pages/ProblemWater.aspx. Accessed September 2017.
• [2] MyCleanDiesel.com. Solution: Water Removal Techniques. Available at http://www.mycleandiesel.com/pages/SolutionWater.aspx. Accessed September 2017.
• [3] ASTM D6304-17. Standard Test Method for Determination of Water in Petroleum Products by Coulometric Karl Fischer Titration. ASTM International, 2017.