Haslinda Zabiri

Artificial Neural Network for Anomalies Detection in Distillation Column

Early detection of anomalies can assist to avoid major losses in term of product degradation, machines’ damages as well as human health issues. This research aims to use Artificial Neural Network to recognize anomalies in the distillation column. The pilot scale distillation column for the ethanol-water system is selected for the study. Faults are generated by variation in feed rate, feed composition and reboiler duty using Aspen Plus® dynamic simulation. The effect of these faults on process variables i.e. changes in distillate and bottom composition, distillate and bottom temperature, bottom flow rate, and the pressure drop is observed. The network is trained using back propagation algorithm to determine root mean square error (RMSE). Based on RMSE minimization, the (6-8-6) net serves as the best choice for the case studied for efficient fault detection. The presented techniques are general in nature and easily applicable to various other industrial problems.

Fault detection in distillation column using NARX neural network

Fault detection in the process industries is one of the most challenging tasks. It requires timely detection of anomalies which are present with noisy measurements of a large number of variable, highly correlated data with complex interactions and fault symptoms. This study proposes the robust fault detection method for the distillation column. Fault detection and diagnosis (FDD) for process monitoring and control has been an effective field of research for two decades. This area has been used widely in sophisticated engineering design applications to ensure the proper functionality and performance diagnosis of advanced and complex technologies. Robust fault detection of the realistic faults in distillation column in dynamic condition has been considered in this study. For early detection of faults, the model is based on nonlinear autoregressive with exogenous input (NARX) network. Tapped delays lines (TDLs) have been used for the input and output sequences. A case study was carried out with three different fault scenarios, i.e., valve sticking at reflux and reboiler, and tray upset. These faults would cause the product degradation. The normal data (no fault) is used for the training of neural network in all three cases. It is shown that the proposed algorithm can be used for the detection of both internal and external faults in the distillation column for dynamic system monitoring and to predict the probability of failure.

A kinetic-based simulation model of palm kernel shell steam gasification in a circulating fluidized bed using Aspen Plus®: A case study

ABSTRACT A detailed simulation model for hydrogen production using catalytic steam gasification of palm kernel shell in an atmospheric dual fluidized bed gasifier using an Aspen Plus® simulator is developed. The catalytic adsorbent-based steam gasification of palm kernel shell is studied in a pilot scale dual fluidized bed reactor using coal bottom ash as a catalyst for hydrogen and syngas production. The use of a catalyst along with the adsorbent improved tar cracking and enhanced the hydrogen content of syngas. The effect of temperature and the steam–biomass ratio on hydrogen yield, syngas composition and lower and higher heating values was studied. An increase in steam–biomass ratio enhanced the hydrogen content from 60 to 72 mol%%. The maximum value of hydrogen production, i.e. 72 vol% was achieved at a steam–biomass ratio of 1.7. The use of adsorbent and coal bottom ash had a significant effect on hydrogen and syngas yield. A maximum of 80.1 vol% hydrogen was achieved at a temperature of 650 °C with a 1.25 steam–biomass ratio and 0.07 wt% coal bottom ash.

Dynamic Modelling for High Pressure CO 2 Absorption from Natural Gas

This paper reports the dynamic simulation model of high content CO2 from natural gas at elevated pressure. The common process of CO2 modelling are mostly reported in steady state condition at atmospheric pressure. However, disturbances such as startup, shut down, and temperature rise might occur during the absorption process. Therefore, the dynamic study has been conducted in this paper via equilibrium approach with some adjustments to observe the efficiency of CO2 removal at the top of the column. Input data for the simulation had been acquired from the pilot plant in Universiti Teknologi PETRONAS (UTP). Aspen Dynamic simulator is not able to support the rate based approach and therefore, several adjustments such as the number of stages and Murphree efficiency need to be imposed on the equilibrium stage model to produce similar result as the pilot plant and as well as rate based approach. The error percentage of CO2 removal observed between actual plant and simulation using equilibrium based approach is less than 5% with several adjustment implemented in the simulator. The results show that the equilibrium approach with some adjustments is able to replicate the pilot plant under dynamic conditions. In dynamic study, the lean solvent flowrate is varied to study the performance of CO2 removal and it is observed the higher solvent lean solvent flowrate improves the efficiency of CO2 removal.

Simulation of CO 2 Rich Natural Gas Pilot Plant Carbon Dioxide Absorption Column at Elevated Pressure Using Equilibrium and Rate Based Method

This paper reports a steady state model for CO2 removal using MEA solvent that operates at elevated pressure and the behaviour that affects the performance of CO2 absorption process. All the input for the simulation has been acquired from the experimental work using pilot plant which is located at Universiti Teknologi PETRONAS (UTP). Steady state simulation has been demonstrated using Aspen Plus utilizing both equilibrium and rate based approaches. Modifications for the equilibrium based method has been done to ensure similarity between rate based and equilibrium based simulation. Since Aspen Dynamic does not support rate based model, adjustment made to the equilibrium model will enable the model to be used for future studies which involves dynamic and control study. The most relevant input parameters of the equilibrium model are methodically varied and the influence of that variation on the simulation results based on CO2 removal percentage was monitored. The evaluation has been conducted to observe the percentage of CO2 removal by setting the Murphree efficiency and varying number of stages of absorber unit.

Aspen Plus® Simulation Studies of Steam Gasification in Fluidized Bed Reactor for Hydrogen Production Using Palm Kernel Shell

In this paper, a steady state simulation for hydrogen production from steam gasification of Palm kernel shell was developed and studied. The gasification pilot plant process has been modelled in Aspen Plus® using Gibbs reactor (R-Gibbs). The effects of different operating parameters using sensitivity analysis, including gasification temperature 600–900 °C and steam flow rate (1 to 2 kg/hr.), on hydrogen yields and Syngas composition were investigated. The simulation results have shown the main gas components in Synthesis gas were H2, CO, CO2, CH4. The product gas hydrogen yield increases with the increase in temperature. The hydrogen concentration improved from 22.52 vol. % to 36.06 vol.%, but the CO concentration decreased from 37.53 vol.% to 28.37% with increasing temperature from 650–900 °C under the operating parameters of the steam flow rate of 1.56 kg/hr.

A Comparison Study between Integrated OBFARX-NN and OBF-NN for Modeling of Nonlinear Systems in Extended Regions of Operation

In this paper the combination of linear and nonlinear models in parallel for nonlinear system identification is investigated. A residuals-based sequential identification algorithm using parallel integration of linear Orthornormal basis filters-Auto regressive with exogenous input (OBFARX) and a nonlinear neural network (NN) models is developed. The model performance is then compared against previously developed parallel OBF-NN model in a nonlinear CSTR case study in extended regions of operation (i.e. extrapolation capability).

Identification of Nonlinear Systems Using Parallel Laguerre-NN Model

In this paper, a nonlinear system identification framework using parallel linear-plus-neural networks model is developed. The framework is established by combining a linear Laguerre filter model and a nonlinear neural networks (NN) model in a parallel structure. The main advantage of the proposed parallel model is that by having a linear model as the backbone of the overall structure, reasonable models will always be obtained. In addition, such structure provides great potential for further study on extrapolation benefits and control. Similar performance of proposed method with other conventional nonlinear models has been observed and reported, indicating the effectiveness of the proposed model in identifying nonlinear systems.