Researchers in Brazil have demonstrated that near-infrared (NIR) spectroscopy, combined with chemometric modeling, can accurately and non-invasively analyze methane, carbon dioxide, and water content in high-pressure natural gas systems.
A team of researchers from Brazilian institutions tested the effectiveness of using near-infrared (NIR) spectroscopy in conducting real-time compositional analysis of natural gas mixtures under high-pressure conditions. This study, which was published in the journal Vibrational Spectroscopy, shows that NIR spectroscopy, when combined with advanced chemometric modeling, can accurately quantify methane (CH₄), carbon dioxide (CO₂), and water vapor in compressed gas systems, even under pressures as high as 120 bar (1).
Currently, the oil and gas industry is undergoing a massive shift. As safety and efficiency standards grow more onerous, oil and gas companies are under pressure to optimize operations to meet these standards. One of the operations that are being increasingly optimized is natural gas dehydration. This process is important because it allows natural gas to be stored and transported safely and efficiently (2). When natural gas is extracted, it is often filled with water vapor. Dehydration removes this water vapor so machinery, including vehicles, can run smoothly without the risk of corrosion (2). Normally, the summer is when most of the gas goes through the dehydration process in preparation for the cold winter months (2).
Unfortunately, current techniques for detecting and quantifying such components often involve complex, time-consuming sampling procedures that can introduce errors or delays. Therefore, in this study, the researchers proposed using NIR spectroscopy for this purpose because it is non-destructive, rapid, and capable of providing accurate measurements in situ, without the need for physical sampling (1).
The researchers used partial least squares (PLS) regression models to interpret the NIR spectral data, achieving high coefficients of determination (R²) of 0.990 for methane, 0.993 for carbon dioxide, and 0.947 for water (1). The corresponding root mean square error of prediction (RMSEP) values, which are 4.43% for CH₄, 3.55% for CO₂, and 8.35% for H₂O, further confirm the technique's predictive strength (1).
Water content, which proved slightly more challenging to quantify, was still measured within acceptable experimental uncertainty, particularly when benchmarked against the Quartz Crystal Microbalance (QCM) method, a well-established reference for moisture quantification (1).
To evaluate the method, the team conducted experiments at room temperature (298.15 K) and pressures up to 120 bar, varying CO₂ concentrations from 0% to 50%. Spectra were collected and processed using PLS modeling, and QCM was used for validation of water measurements in binary and ternary gas mixtures, including CH₄ + H₂O, CO₂ + H₂O, and CH₄ + CO₂ + H₂O (1).
There are several key takeaways and implications of this study. First, the method presented here allows the user to monitor gas composition and moisture in one step without needing to extract physical samples. As a result, this has the potential to improve process control in the natural gas sector (1). Real-time monitoring allows for more responsive adjustments to operations, reducing the risk of system failures and optimizing the energy efficiency of dehydration and separation processes (1).
Moreover, the researchers point out that this technique could be just as valuable in the realm of Carbon Capture and Storage (CCS), where CO₂ is compressed and transported under high pressures. In these systems, monitoring trace amounts of water is equally vital to prevent corrosion and ensure long-term integrity of storage facilities (1).
Another important takeaway is that NIR systems have shown to be sustainable alternatives to traditional methods. By improving the precision of dehydration and separation processes, NIR systems can help minimize energy use and reduce emissions associated with gas processing (1).
The authors noted in their study that future work should focus on expanding the model’s robustness across broader temperature and compositional ranges and integrating the system into field-scale pilot programs (1). This study presents a new tool for the energy sector that can lead to more sustainable, data-driven approaches to gas processing in a high-pressure world.
