The Chinese Journal of Process Engineering

Select

A robust method for chemical process monitoring based on Johnson transformation

Ji WANG Nan LIU Minggang HU Wende TIAN

The Chinese Journal of Process Engineering 2021, 21 (12): 1491-1502. DOI: 10.12034/j.issn.1009-606X.220340

Abstract （202）

PDF （2151KB）（29）

Save

A tiny fault in chemical plants is likely to cause an enormous accident possibly with heavy losses of personnel, property, and environment. Therefore, process monitoring is demanded to timely detect faults and identify fault variables, so as to avoid deterioration of tiny faults into accidents. Nowadays principal component analysis (PCA) is the most widely used method in chemical process monitoring practice with its simplicity and effectiveness. However, it has some drawbacks. First, it roughly assumes process data as Gaussian distribution. But sometimes it is not satisfied. Furthermore, PCA uses variance and covariance (also called Pearson correlation coefficients) as criterion to choose principal components, however they are not robust in capturing nonlinear data variation. To alleviate these problems, an improved PCA-a Johnson transformation based robust method for process monitoring (JSPCA) was proposed in this work. First, Johnson transformation was introduced to make process data obey Gaussian distribution. Second, the Spearman correlation coefficient matrix instead of covariance matrix was established to extract principal components and span feature space. Finally, process data were projected into feature space where T2 and SPE statistics were obtained for process monitoring. The proposed method had its fault detection ability tested in the benchmark TE process with comparison of PCA and KPCA. The results showed that the proposed method had higher fault detection rates than PCA and KPCA when using T2 as detection indicator. However, the proposed method with SPE as detection indicator had higher false alarm rates than PCA and KPCA. As for fault diagnosis ability, the proposed method was tested against fault 5 and fault 10 of TE process and diagnoses fault variables more precisely than PCA and KPCA. The proposed method was better than PCA and KPCA and it was worth promoting.

Related Articles | Metrics

Select

Modeling and analysis of liquid real-time continuous leakage in horizontal liquid ammonia tank

Juanxia HE Dongmei ZHOU Lei LIU Qiyong ZHOU Liwen HUANG Jianting YAO

The Chinese Journal of Process Engineering 2021, 21 (6): 731-740. DOI: 10.12034/j.issn.1009-606X.220171

Abstract （343）

PDF （492KB）（99）

Save

Based on Van der Waals equation and theory of fluid mechanics, the liquid real-time continuous leakage model of horizontal liquid ammonia tank was established considering the changes of tank pressure and liquid surface area. Mathematical modeling of a horizontal liquid ammonia tank in a refrigeration company was performed by this model, and the calculation results were compared with PHAST simulation results. The results showed that the decreasing range of liquid level height h grew slowly and then increased quickly, the decreasing range of liquid leakage mass flow rate Qm and liquid leakage rate v and the increasing range of liquid leakage mass m decrease slowly. At the beginning of leakage, Qm and v were the maximum values, at the end of leakage, m was the maximum values. When the diameters of leakage hole were 5, 30 and 100 mm, leakage time t were 29884.027, 837.289, 77.550 s, Qm(max) were 0.552, 19.913 and 221.160 kg/s, v (max) was 46.733 m/s, m(max) were 10255.649, 10339.923 and 10572.760 kg, respectively. The deviation between the calculation results of the model and the PHAST simulation results was less than 24%. From the analysis of parameter variation and risk emergency, the model had the certain applicability for the theoretical calculation of liquid leakage in horizontal liquid ammonia tank.

Related Articles | Metrics

Select

Fault diagnosis for chemical processes based on deep residual network

Lusheng ZHONG Xiangming XIA

Chin. J. Process Eng. 2020, 20 (12): 1483-1490. DOI: 10.12034/j.issn.1009-606X.219374

Abstract （421）

PDF （987KB）（220）

Save

A fault diagnosis method for chemical processes based on deep residual network (DRN) was proposed, which could automatically extract fault features from a large number of chemical processes operation data. The model adopted the shortcut connections to alleviate the training difficulty in the traditional deep neural network, and adopted the batch normalization (BN) method, which could effectively alleviate the problem of vanishing/exploding gradients. The Tennessee Eastman (TE) process was used as the experimental object to evaluate the diagnostic performance of the proposed method. The proposed method and the previous TE process fault diagnosis method based on traditional deep learning model were compared. Furthermore, the effects of the number of layers, BN technology and residual structure on fault diagnosis rate were studied. Finally, the output of some layers was visualized by the t-distributed stochastic neighbor embedding (t-SNE) method. The results showed that the model achieved an average fault diagnosis rate of 94% and an average false positive rate of 0.30% for 21 working conditions, showing more excellent diagnostic performance. The two-dimensional scatter plot of the output layer showed clear clustering, which indicated that the proposed DRN model can accurately diagnose the faults.

Related Articles | Metrics

Select

Data analysis and optimization of butane oxidation reactor

Mingyu CHEN Zheli WEI Jian LI Xiangdong ZHU Ruhui YANG Xing XIANG Erqiang WANG Xiaoxiang SUN

Chin. J. Process Eng. 2020, 20 (7): 870-876. DOI: 10.12034/j.issn.1009-606X.219321

Abstract （500）

PDF （445KB）（180）

Save

As an important reaction system, butane oxidation to produce maleic anhydride has already been industrialized with lots of advantages compared with other production processes. In this process, fixed-bed tubular reactor was used and recycling molten salt was selected as the cooling media to take out huge amount of reaction heat. Due to the complexity of the reactor internal structure and the reaction mechanism, it is difficult to develop a rigorous mathematical model to simulate and optimize this reactor. Black-box models, such as artificial neural network (ANN), could not provide detailed information about inherent mechanism of research process, and could be only used in the manner of interpolation within fixed range. The principle component analysis (PCA) is one of the most popular statistical methods for data mining and analysis. PCA can help to reduce the dimensionality of the variable space by representing it with a few orthogonal (uncorrelated) variables that capture most of its variability. So PCA retains those characteristics of the data set that contribute most to its variance, by keeping lower-order principal components (the ones that explain a large part of the variance present in the data) and ignoring higher-order ones (that do not explain much of the variance present in the data). In this work, lots of historical data of butane oxidation reactor was firstly selected from the DCS device, and then corrected to be as the basis of data mining analysis. The PCA technology was used to dig the relationship between these reactor parameters. The results showed that these outliers can be effectively detected as abnormal or normal data point and the former data would be removed from the data before next analysis. It was also found that there was a negative correlation between the conversion of butane and CO/CO2 ratio at reactor outlet. So, these conclusions from this PCA analysis could be used as useful guide for reactor operation and optimization. The principle component analysis (PCA) is one of the most popular statistical methods for data mining and analysis. PCA can help to reduce the dimensionality of the variable space by representing it with a few orthogonal (uncorrelated) variables that capture most of its variability. So PCA retains those characteristics of the data set that contribute most to its variance, by keeping lower-order principal components (the ones that explain a large part of the variance present in the data) and ignoring higher-order ones (that do not explain much of the variance present in the data). In this article, lots of historical data of butane-oxidation reactor was firstly selected from the DCS device, and then corrected to be as the basis of data mining analysis. The PCA technology was used to dig the relationship between these reactor parameters. The results showed that these outliers can be effectively detected as abnormal or normal data point and the former data would be removed from the data before next analysis. It was also found that there was a negative correlation between the conversion of butane and CO/CO2 ratio at reactor outlet. So, these conclusions from this PCA analysis could be used as useful guide for reactor operation and optimization.

Related Articles | Metrics

Select

Process integration and energy analysis of bischofite producing metal magnesium

Zhaoyuan WAN Huan ZHOU

Chin. J. Process Eng. 2020, 20 (5): 609-618. DOI: 10.12034/j.issn.1009-606X.219259

Abstract （511）

PDF （628KB）（176）

Save

Magnesium metal production from bischofite is a process of high energy consumption. It is necessary to explore the process of lowest energy consumption. In this work, with anhydrous magnesium chloride and magnesium oxide as intermediate products, electrolysis and thermal reduction as key methods, a comprehensive process network from bischofite to magnesium metal was constructed, which involved 24 species, 20 chemical processes and 25 process routes. Furthermore, one minimum energy consumption model was proposed to evaluate the thermal effect of multi–chemical process, or multi–process routes. Using the standard enthalpy of formation and temperature-depended isobaric molar heat capacity, the energy consumption and heat removal of all 25 process routes were calculated. The results showed that the optimum path based on thermal reduction method was converting bischofite to magnesium hydroxide by lime method, calcining to obtain magnesium oxide, and further reducing to magnesium metal by aluminum. The energy consumption was 360.15 kJ/mol, and the heat released was –315.46 kJ/mol. Compared with this, the better path of electrolysis was producing magnesium hydroxide by lime method, calcining to get magnesium oxide and further producing magnesium metal by electrolyze in molten electrolyte. The energy consumption of the process was 738.54 kJ/mol, and the heat released was –135.42 kJ/mol. Because of the high energy consumption of anhydrous magnesium chloride preparation, it was not in the optimal path.

Related Articles | Metrics

Content of Process System Integration & Chemical Safety in our journal