Experimental dataset for system identification of a dual-phase gas-liquid transmembrane pressure system.

The data presented in this article describe the dynamic input-output responses of a dual-phase gas-liquid transmembrane pressure system operating under different water flow rates and actuator excitations. The dataset is acquired from a laboratory-scale experimental platform consisting of interconnected CO₂ gas and water channels separated by a porous polytetrafluoroethylene (PTFE) membrane module. The gas and water control valves are used to regulate the gas and liquid flow paths, while the liquid channel is supplied by a controlled water pump. Gas inlet pressure and water outlet pressure are measured using industrial pressure transducers and synchronously acquired by a Siemens S7-1200 programmable logic controller at a sampling frequency of 1 Hz. Data are collected under three water flow-rate conditions. For each condition, three open-loop excitation profiles-sine-wave, sawtooth, and multi-step commands-are applied to gas and water control valves. Each experiment lasts 20 min and generates 1200 samples. The complete dataset consists of nine XLSX files containing 10,800 synchronized samples. Each file contains five channels: time, commanded gas-valve position, commanded water-valve position, gas inlet pressure, and water outlet pressure, with clearly labeled variables and physical units. This dataset can be reused to analyze the system responses under different excitation profiles and water flow-rate conditions, characterize transient and coupled input-output behavior, as well as develop and validate linear and nonlinear multiple-input multiple-output system identification methods. It can also support data-driven dynamic modelling, as well as the subsequent design and evaluation of model-based control strategies for gas-liquid transmembrane pressure systems.