RESUMO
Lubricants are important fluids and are commonly used to suppress friction between two metallic surfaces and as a medium for heat transportation. In an industrial plant considered in this study, the base oil mode changes can only be detected based on the kinematic viscosity values obtained using lab analysis. Since the lab analysis data are only available every 8 h, detecting the change in the production modes for 4, 6, and 10 cSt and the transitions among them are significantly delayed, causing unnecessary off-spec products that have to be directed to the slopping tank. In this paper, the innovativeness of the work comes from the idea of trying to unravel the underlying pattern of the plant data that correlate to the changes in the base oil modes and using that to classify hourly the kinematic viscosity values. Hence, a novel industrial application is presented to predict the class of base oil mode change on an hourly basis that can significantly reduce the losses in terms of off spec products and sloping tank wastes. The modes are segregated into three classes based on the values of kinematic viscosity. The classes are C-1 (4 cSt), C-2 (6 cSt), and C-3 (10 cSt). Anything in between the stipulated thresholds is called transition [T-12 (C-1 to C-2), T-21(C-2 to C-1), T-23 (C-2 to C-3), T-31 (C-3 to C-1), and T-32 (C-3 to C-2)]. To unravel the pattern, principal component analysis (PCA) is utilized on 42,000 operating plant data. After a thorough analysis, the third principal component provides the highest correlation to the eight classes of the base oil mode changes [C-1 (4 cSt), C-2 (6 cSt), and C-3 (10 cSt) and the transitions T-12 (C-1 to C-2), T-21(C-2 to C-1), T-23 (C-2 to C-3), T-31 (C-3 to C-1), and T-32 (C-3 to C-2)]. This third principal component is then utilized together with plant process variable values as inputs to four machine learning models, namely, XGBOOST, Random Forest, and CatBoost algorithms to predict the mode of the base oil hourly. The overall comparison analysis shows that utilizing the XGBoost algorithm for the prediction of the eight classes of the base oil modes at a faster hourly rate results in the most consistent classification accuracy of 92.96% for the test set and 89.22% in the deployment set. This capability to predict the mode change in the hourly basis can significantly reduce the losses in terms of off spec products in the production line.
RESUMO
The conversion of biomass through thermochemical processes has emerged as a promising approach to meet the demand for alternative renewable fuels. However, these processes are complex, labor-intensive, and time-consuming. To optimize the performance and productivity of these processes, modeling strategies have been developed, with steady-state modeling being the most commonly used approach. However, for precision in biomass gasification, dynamic modeling and control are necessary. Despite efforts to improve modeling accuracy, deviations between experimental and modeling results remain significant due to the steady-state condition assumption. This paper emphasizes the importance of using Aspen Plus® to conduct dynamics and control studies of biomass gasification processes using different feedstocks. As Aspen Plus® is comprising of its Aspen Dynamics environment which provides a valuable tool that can capture the complex interactions between factors that influence gasification performance. It has been widely used in various sectors to simulate chemical processes. This review examines the steady-state and dynamic modeling and control investigations of the gasification process using Aspen Plus®. The software enables the development of dynamic and steady-state models for the gasification process and facilitates the optimization of process parameters by simulating various scenarios. Furthermore, this paper highlights the importance of different control strategies employed in biomass gasification, utilizing various models and software, including the limited review available on model predictive controller, a multivariable MIMO controller.
RESUMO
An acid gas removal unit (AGRU) in a natural gas processing plant is designed specifically to remove acidic components, such as carbon dioxide (CO2) and hydrogen sulfide (H2S), from the natural gas. The occurrence of faults, such as foaming, and to a lesser extent, damaged trays and fouling, in AGRUs is a commonly encountered problem; however, they are the least studied in the open literature. Hence, in this paper, shallow and deep sparse autoencoders with SoftMax layers are investigated to facilitate early detection of these three faults before any significant financial loss. The dynamic behavior of process variables in AGRUs in the presence of fault conditions was simulated using Aspen HYSYS Dynamics. The simulated data were used to compare five closely related fault diagnostic models, i.e., a model based on principal component analysis, a shallow sparse autoencoder without fine-tuning, a shallow sparse autoencoder with fine-tuning, a deep sparse autoencoder without fine-tuning, and a deep sparse autoencoder with fine-tuning. All models could distinguish reasonably well between the different fault conditions. The deep sparse autoencoder with fine-tuning was best able to do so with very high accuracy. Visualization of the autoencoder features yielded further insight into the performance of the models, as well as the dynamic behavior of the AGRU. Foaming was relatively difficult to distinguish from normal operating conditions. The features obtained from the fine-tuned deep autoencoder in particular can be used to construct bivariate scatter plots as a basis for automatic monitoring of the process.
RESUMO
Promoted potassium carbonate with glycine has been actively investigated as a chemical solvent for the removal of CO2. Though a vast number of studies have been reported for potassium carbonate, dynamic studies regarding this promoted solvent are not yet extensively reported in the literature. In this work, a steady-state simulation has been performed via an equilibrium stage model in Aspen Plus V10 using the experimental data of an absorber from the bench scale pilot plant (MINI CHAS) located in Universiti Teknologi PETRONAS. In this study, 15 wt % K2CO3 + 3 wt % glycine is utilized as the medium for absorption, and the operating pressure is set at 40 bar to imitate the natural gas treatment process. The removal observed from the pilot plant is about 75% and the steady-state simulation with a tuned vaporization efficiency managed to replicate a similar result. The transient analysis is performed via activating a flow-driven method, and the simulation is transferred into Aspen Dynamic. A simple control strategy using a proportional-integral (PI) controller is installed at the gas outlet to monitor the CO2 composition, and further disturbances are introduced at the inlet gas flow rate using a step test and ramp test. The controller is tuned using a trial-and-error method and a satisfactory response is achieved under varying changes in the inlet gas flow rate. Further investigation is carried out using the model predictive controller (MPC), in which 5000 data points are generated through pseudorandom binary sequence (PRBS) analysis for state-space model system identification. The state-space model identified as the best is then used to design the MPC controller. A disturbance rejection test on the MPC controller is conducted via changing the gas flow rate at 5% and a quick response is observed. In conclusion, both MPC and PI controllers managed to produce a good response once the disturbance was introduced within the CO2-potassium carbonate-glycine (PCGLY) system.
RESUMO
The purpose of this paper is to investigate the possible implementation of the Fast model predictive control (MPC) scheme for chemical systems. Due to the difficulties associated with complicated dynamic behavior and model sensitivity, which results in considerable offsets, the Fast MPC controller has not been implemented on the CO2 capture plant based on the absorption/stripping system. The main objective of this work is to evaluate the most appropriate model for implementing the Fast MPC control strategy, which results in fast output responses, negligible offsets, and minimum errors. The steady-state and dynamic simulation models of the CO2 capture plant are designed in Aspen PLUS. In the System Identification Toolbox, multiple state-space models are identified to achieve a highly accurate model for the Fast MPC controller. The Fast MPC controller is then implemented to evaluate the performance under a setpoint tracking mode with ±5 and ±15% step changes. The results showed that the Fast MPC based on the state-space prediction focus model has on average 7.9 times lower offset than the simulation focus model and 10.4 times lower integral absolute error values. The comparison study concluded that the Fast MPC control strategy performs efficiently using prediction-based focus state-space models for CO2 capture plants using the absorption/stripping system with minimum offsets and errors.
RESUMO
The chemical process industry has become the backbone of the global economy. The complexities of chemical process systems have been increased in the last two decades due to online sensor technology, plant-wide automation, and computerized measurement devices. Principal component analysis (PCA) and signed directed graph (SDG) are some of the quantitative and qualitative monitoring techniques that have been widely applied for chemical fault detection and diagnosis (FDD). The conventional PCA-SDG algorithm is a single-scale FDD representation origin, which cannot effectively solve multiple FDD representation origins. The multiscale PCA-SDG wavelet-based monitoring technique has potential because it easily distinguishes between deterministic and stochastic characteristics. This study uses multiscale PCA-SDG to detect, diagnose the root cause and identify the fault propagation path. The proposed method is applied to a continuous stirred tank reactor system to validate its effectiveness. The propagation route of most process failures is detected, identified, and diagnosed, which is well-aligned with the fault description, demonstrating a satisfactory performance of the suggested technique for monitoring the process failures.