PHASE TRANSITION DETECTION IN CO₂–OIL LIGHT FRACTION SYSTEMS USING SWIR-BASED PIXEL INTENSITY ANALYSIS

Phase-transition detection in complex fluid mixtures remains challenging, especially in opaque or semi-transparent systems where visual methods are subjective and often limited. Several techniques have been employed for this purpose, including calorimetry, rheological measurements, densimetry, and light scattering. High-pressure PVT cells offer a unique advantage by allowing multiple sensing approaches to be applied simultaneously under controlled conditions, enabling cross-validation of phase-equilibrium indicators. Within this context, this work investigates the use of Mean Pixel Intensity (MPI), derived from SWIR camera monitoring, as a quantitative optical metric for phase-transition detection.

Experiments were conducted with a mixture of dead oil light fraction and CO₂ subjected to Static-Sinthetic Method in a PVT cell at an isotherm state. SWIR video data were acquired by a LabVIEW supervisory system and stored as average grayscale values (14-bit) for each pressure step. A Python-based workflow was developed using pandas, numpy, and matplotlib to process the raw pressure, transmittance, volume, and MPI data. Saturation pressure was determined using the tangent method applied over all measured variables.

At bubble-point conditions, MPI–pressure curves exhibited clear inflection behavior, yielding saturation pressures in close agreement with conventional transmittance measurements, PV curves across the investigated temperature range. At dew-point conditions, MPI produced a response qualitatively coherent with visual observation, whereas the PV curve showed no variation, not being suitable for transition detection. Overall, the MPI technic shows hight compatibility with those of well stablish technics, presenting promising potential as a complementary, automated optical tool for phase-transition detection in high-pressure PVT studies, such as crude oil mixtures.