A Novel Approach for Prediction of Industrial Catalyst Deactivation Using Soft Sensor Modeling

Abstract

Soft sensors are used for fault detection and prediction of the process variables in chemical processing units, for which the online measurement is difficult. The present study addresses soft sensor design and identification for deactivation of zeolite catalyst in an industrial-scale fixed bed reactor based on the process data. The two main reactions are disproportionation (DP) and transalkylation (TA), which change toluene and C 9 aromatics into xylenes and benzene. Two models are considered based on the mass conservation around the reactor. The model parameters are estimated by data-based modeling (DBM) philosophy and state dependent parameter (SDP) method. In the SDP method, the parameters are assumed to be a function of the system states. The results show that the catalyst activity during the period under study has approximately a monotonic trend. Identification of the system clearly shows that the xylene concentration has a determining role in the conversion of reactions. The activation energies for both DP and TA reactions are found to be 43.8 and 18 kJ/mol, respectively. The model prediction is in good agreement with the observed industrial data.

Random Forest Regression to Predict Catalyst Deactivation in Industrial Catalytic Process

“…A combination of laboratory measurements and online analyzers was used to monitor the decrease in catalyst activity [8]. Others used soft sensors to identify variations in catalyst activity using the data-based modeling (DBM) philosophy [9].…”

Soft sensor hybrid model of dynamic liquid level for sucker rod pump oil wells

“…The error between the dynamic liquid level and the measured value is calculated by the data-driven model, which makes the hybrid model have better performance to approximate the real dynamic liquid level value. Hybrid modelling technology and data-driven soft sensing have been successfully applied in many complex industrial processes: chemical processes in the work by Dang et al (2020), Gharehbaghi and Sadeghi (2016), Mecchia et al (2019), Sun et al (2019aSun et al ( , 2019b, and Tian et al (2016); metallurgical and meteorological processes in the work by Chen et al (2017), Saqib and Porter (2016), Tian (2020b), Wang et al (2019), Saqib et al (2016), and Wang et al (2019); and oil and gas production processes in the work by Li XY et al, (2020), Li K et al, (2020a), Li K et al, (2020b), Tian ZD (2020a.…”

Modeling of a magneto-rheological damper: An improved multi-state-dependent parameter estimation approach

“…The SDP estimation method has been used in various fields (Bidar et al, 2017; Gharehbaghi and Sadeghi, 2016; Li et al, 2014, 2016; Li and Lu, 2013; Ratto and Pagano, 2010; Taylor and Robertson, 2013). The general concept of SDP modeling involves multi-state dependency, in which each parameter can be a function of several state variables.…”