Data-Model Fusion-Driven Method for Fault Quantitative Diagnosis of Heat Exchanger

Abstract

Heat exchangers play essential roles in the oil and gas production process for convective heat transfer and heat conduction. The health management of heat exchangers stays in the direct monitoring of performance parameters. Aiming at the difficulty of precise fault identification and quantification for heat exchangers in multiple unknown failure modes, a data-model fusion-driven fault quantitative diagnosis method is proposed. Firstly, based on the monitoring data such as temperature, pressure and flow rate, the secondary parameters characterizing the heat exchanger running state are constructed combined with structural physical parameters. Then, by analyzing the correlation among parameter variation, failure modes and deterioration degree, a qualitative inference model of heat exchanger is formed for fault identification, where weights of parameters are introduced based on their sensitivity for different failure modes. After the fault mode is identified, to achieve quantitative analysis of the failure degree, an index-integrated mechanism equation is constructed using monitoring data and secondary parameters, where the index is dynamically modified by online data. Finally, a heat exchanger experiment is carried out to demonstrate the robustness and accuracy of the proposed method.