Inline process monitoring of the moisture content in propylene oxide

Importance of the Topic

Propylene oxide is a key industrial intermediate with a global output exceeding 10 million tonnes annually. Accurate control of residual water in PO streams is critical to prevent undesirable side reactions, maintain high product quality, and ensure operational safety. Real-time inline moisture monitoring minimizes manual handling of this hazardous, hygroscopic chemical and delivers rapid feedback for process optimization.

Objectives and Study Overview

This study outlines the deployment of a near-infrared spectroscopy (NIRS) system for continuous, inline measurement of low ppm-level water in PO. The aim is to integrate an explosion-proof analyzer capable of monitoring up to nine process points simultaneously, offering rapid detection of moisture deviations and enhancing overall process control.

Methodology and Instrumentation

Measurements were performed in the 1850–1950 nm spectral range using transmission-mode stainless steel flow cells installed directly in the PO process line. A multiplexed NIRS XDS Process Analyzer with single-fiber optics enabled connection of up to nine probes located at distances up to 100 m from the central unit. Calibration models were built by correlating inline spectra with Karl Fischer titration data.

Used Instrumentation

• NIRS Analyzer: XDS Process Analyzer SingleFiber in explosion-proof housing
• Flow Cells: SS316 fixed pathlength (2 mm), adjustable pathlength (0.5–12 mm), and PTFE flow-through options
• Optical Probes: Single-fiber configuration with up to nine channels
• Reference Method: Karl Fischer titration for calibration and validation

Main Results and Discussion

The NIRS calibration model demonstrated reliable prediction of water concentrations in the typical ASTM range (20–30 mg/L). Online predictions were initially lower than offline titrations due to atmospheric moisture uptake during sample handling. Introducing inline KF titration improved model accuracy, ensuring that spectroscopic measurements closely reflected true process conditions. Continuous monitoring across multiple points allowed prompt detection of moisture spikes and rapid corrective actions.

Benefits and Practical Applications of the Method

• Real-time control reduces reaction inefficiencies and increases yield.
• Eliminates manual sampling, lowering operator exposure to hazardous chemicals.
• Accelerates response to process anomalies, minimizing downtime and waste.
• Delivers fast ROI through improved consistency and reduced off-spec material.

Future Trends and Potential Applications

Integration of inline NIRS with advanced data analytics and process automation will further enhance predictive control and maintenance. Extending this approach to other moisture-sensitive streams, such as naphtha fractions or peroxide processes, can leverage existing hardware. Emerging machine learning algorithms may refine calibration models, boosting accuracy and enabling proactive process adjustments.

Conclusion

Inline NIRS offers a safe, fast, and robust solution for monitoring low-level moisture in propylene oxide. Coupled with periodic Karl Fischer titration for model validation, this approach improves safety, product quality, and operational efficiency in demanding industrial settings.

References

• Kawabata T, Yamamoto J, Koike H, Yoshida S. Trends and Views in the Development of Technologies for Propylene Oxide Production. Sumitomo Kagaku. 2019; pp 4–11.
• Kleimeier. Nahinfrarotspektroskopie Produktionsprozesse Unter Der Lupe. GIT Labor-Fachzeitschrift. 2018; 36–38.