加氢裂化动力学模型及其工业应用

在文献调查及对加氢加过程深入研究的基础上，建立了稳态的加氢裂化反应动力学模型，并将该模型较成功地应用于工业加氢氧化反应过程的模拟。     

氢分压对加氢裂化过程的影响

分析了氢分压加氢裂化过程各种反应的影响以及加氢裂化过程中氢分压对裂化转化率，产品分布及产品性质的综合影响，提出了后的加氢裂化过程对催化剂和工艺的一些基本要求。     

加氢裂化产品的重新分割

加氢裂化产品性质优良 ,能直接用作商品油或后续加工 ,该工艺过程又很灵活 ,可以随市场变化和季节的不同而改变生产方案 ,以获得较高的经济效益。文中提出的方法可以根据现有生产过程 ,预测产品重新分割后产品产率和性质 ,以进行生产调优 ,满足市场需求。1　加氢裂化产品的重新分割基于加氢裂化的广泛应用 ,提出一种新的预测分割模型 ,该模型以实沸点曲线的切割为理论依据 ,输入某种原料经加氢裂化以后所得的产品的相对密度、产率及恩氏蒸馏温度 ,重新划分产品 ,进而预测重新划分产品的性质 ,为不同的生产方案提供依据。以加氢裂化产品基础…     

加氢裂化集总动力学模型的研究Ⅱ．正十烷加氢裂化集总模型

选用３８２４加氢裂化催化剂，对正十烷加氢裂化动力学进行了研究。根据Ｗｅｅｋｍａｎ集总理论，建立了正十烷加氢裂化四集总动力学模型。用Ｍａｒｑｕａｒｄｔ法估计了各反应速率常数，确定了较完善的速率表达式和表现活化能，同时讨论了空速、温度、压力和反应活化能对产物分布的影响，为石油馏份加氢裂化集总动力学研究提供了基础数据。     

加氢裂化集总反应动力学模型研究

根据加氢裂化反应机理和反应特征，建立了通用性较强的加氢裂化集总动力学模型；采用序列二次规划方法进行参数估计，确定了反应动力学参数。研究结果表明，模型能很好地拟合实验数据，并具有良好的预测性能。     

加氢裂化常压塔轻石脑油干点软测量

先进控制技术可以把若干个变量控制在期望的控制点上，从而达到装鼍效益的整体优化。近年来，各种先进控制软件在催化裂化装置、加氢裂化等装置上得到了广泛的应用。为了实施加氢裂化装置的先进控制，需要在线计算其产品质量指标。轻石脑油干点就是其中一个质量指标。为了避免单纯依靠过程数据进行软测量建模而导致的计算结果与工艺分析结果相矛盾的情况，结合过程数据和质量指标的关联公式，利用粒子群算法建立轻石脑油的软测量模型。该软测量模型已在现场实施，应用结果表明，该模型有较高精度。     

加氢裂化技术的新进展

加氢裂化是重要的现代炼油工艺，我国现有加氢裂化装置加工能力为９８６万ｔ／ａ，今后还将进一步发展。近１０几年来，我国不仅自行设计，建造大型加氢裂化装置，同时抚顺石油化工研究院还根据需要发展了缓和加氢裂化、中压加氢裂化、中压加氢改质和润滑油加氢处理等新工艺，     

中低温煤焦油加氢裂化集总动力学模型的研究

以煤焦油为原料,在高压固定滴流床反应器中,以工业NiMo/Al2O3为催化剂,考察了360-380℃范围内煤焦油的产物分布,基于此建立了5集总煤焦油加氢裂化动力学模型。动力学模型的集总包括：未反应的煤焦油、柴油、汽油、气体和焦炭。通过对实验产物与模型预测产物的对比数据,发现本文所建立的动力学模型可以用于煤焦油加氢裂化过程。同时,基于动力学模型,进一步分析了煤焦油的加氢裂化机理：在整个煤焦油加氢裂化过程中,柴油馏分可作为反应中间组分。     

重质油加氢裂化反应动力学的研究

本文依据大量文献资料，介绍了重质油加氢裂化反应动力学的研究现状，对各种动力学模型的优缺点进行了评论，指出了催化裂化和加氢裂化反应动力学的相似性和进一步深化研究的方向。     

用正交配置法模拟加氢裂化反应器

结合物料平衡、能量平衡和加氢裂化反应动力学方程,按照加氢裂化反应的特点,建立了加氢裂化反应器6集总动态机理模型。反应动力学采用原料油、柴油、航空煤油、重石脑油、轻石脑油、气体6集总模型。对总动态机理模型提出了正交配置法和龙格-库塔法相结合的求解方法。新的求解方法计算量小、精度高,并且能求得微分数学模型的近似解析解。应用6集总动态机理模型对加氢裂化反应器进行了模拟,考察了模型的预测精度。模拟结果表明,该模型能较好地模拟和预测加氢裂化产品的收率分布和反应器的温度分布,具有较高的预测精度,模型可靠。     

A Kinetic Model for Hydrocracking Process

ＡＫｉｎｅｔｉｃＭｏｄｅｌｆｏｒＨｙｄｒｏｃｒａｃｋｉｎｇＰｒｏｃｅｓＷｕＸｉｚｈｏｎｇ，ＬｉＣｈｅｎｇｌｉｅ，ＦａｎｇＸｉａｎｇｃｈｅｎ，ＣｈａｎｇＺｈｅｎｍｉｎｇ（ＥａｓｔＣｈｉｎａＵｎｉｖｅｒｓｉｔｙｏｆＳｃｉｅｎｃｅ＆Ｔｅｃｈｎｏｌｏｇｙ，Ｓ...     

A Four Lump Kinetic Model for the Simulation of the Hydrocracking Process

Hydrocracking is a very important secondary refining process used to convert low value vacuum gas oils into high value fuels. The chemistry of the hydrocracking process is very complex due to the involvement of high molecular weight complex hydrocarbons in the reactions. Process modeling and simulation of the hydrocracking unit is very challenging due to the complexities of chemistry and the process. In the present work, a mathematical model is described to analyze the performance of the hydrocracking process in terms of product yields. A four lump discrete lumping approach is employed with a hydrocracking reaction scheme involving six reactions. The kinetic constants for the reactions were estimated by minimizing the error between the experimental and predicted yields of kinetic lumps. Experimental data reported by Ali et al. (2002) were used to validate the proposed model. The model predictions were found to agree well with the experimental data. The proposed model can be used to simulate the performance of commercial hydrocrackers using kinetic parameters estimated from pilot plant experiments.     

HYDROZEN TRANSFER HYDROCRAKING OF C.PROCERA UNDER AMBIENT PRESSURE CONDITIONS TO GET VALUE ADDED CHEMICALS AND FUELS

Biomass is renewable source of energy while the reserves of petroleum arc being depleted. The latex of a potential petrocrop, Colotropis procera, a lalicifcr, arid-plant which is rich in hydrocarbon type triterpene compounds etc. was found lo be a better feed slock for thermal hydrocracking as compared to whole plant biomass inlcrms of liquid product yield. Studies of chemical reaction dynamics of the thermal cracking of latex at 200-400°C showed that the process should be termed as hydrogen-tranfer (H-T) hydrocracking of latex under ambient pressure conditions. The hydrogen rich cracked trilcrpenoids act as the H-donors in this process, where nascent hydrogen atoms and free radicals chemically plug the cracked moities to stabilise these. Latex was also coagulated and the H-T hydrocracking of the feedstock coagulum gave a higher yield of cracked oil in comparision lo that from the dried latex. A model triterpene compound, ursolic acid has been subjected to H-T hydrocracking to understand the process of hydrocracking of latex under similar conditions and it was found that triterpencs on H-T hydrocracking produced only liquid and gaseous products and no solid char. The temperature for hydrocracking of latex has been optimized to 350°C and molecular sieve was round to catalyse the H-T Hytrocraking process to yield more liquid product The distillation range of cracked latex on(CLO)Obtained from H.T Hytrocracking of C procera Latex indicated that it can be used as fuel. Moreover CLO resembled diesel fuels and was predominantly paraffinic in nature as characterised by NMR and FTIR spectral analysis. A process has been recommended for gelling value added fuels and chemicals from C. procera latex.     

HYDROZEN TRANSFER HYDROCRAKING OF C.PROCERA UNDER AMBIENT PRESSURE CONDITIONS TO GET VALUE ADDED CHEMICALS AND FUELS

ABSTRACT

Biomass is renewable source of energy while the reserves of petroleum arc being depleted. The latex of a potential petrocrop, Colotropis procera, a lalicifcr, arid-plant which is rich in hydrocarbon type triterpene compounds etc. was found lo be a better feed slock for thermal hydrocracking as compared to whole plant biomass inlcrms of liquid product yield. Studies of chemical reaction dynamics of the thermal cracking of latex at 200-400°C showed that the process should be termed as hydrogen-tranfer (H-T) hydrocracking of latex under ambient pressure conditions. The hydrogen rich cracked trilcrpenoids act as the H-donors in this process, where nascent hydrogen atoms and free radicals chemically plug the cracked moities to stabilise these. Latex was also coagulated and the H-T hydrocracking of the feedstock coagulum gave a higher yield of cracked oil in comparision lo that from the dried latex. A model triterpene compound, ursolic acid has been subjected to H-T hydrocracking to understand the process of hydrocracking of latex under similar conditions and it was found that triterpencs on H-T hydrocracking produced only liquid and gaseous products and no solid char. The temperature for hydrocracking of latex has been optimized to 350°C and molecular sieve was round to catalyse the H-T Hytrocraking process to yield more liquid product The distillation range of cracked latex on(CLO)Obtained from H.T Hytrocracking of C procera Latex indicated that it can be used as fuel. Moreover CLO resembled diesel fuels and was predominantly paraffinic in nature as characterised by NMR and FTIR spectral analysis. A process has been recommended for gelling value added fuels and chemicals from C. procera latex.     

Kinetic model for hydrocracking of Iranian heavy crude with dispersed catalysts in slurry-phase

A five-lump kinetic model has been presented to describe the process of heavy oil hydrocracking, and tested in moderate condition. In the condition of low temperature and pressure relatively, coke formation can be controlled and the model is well applied with product distribution. In this paper, we explored the process of hydrocracking in slurry-phase by using dispersed catalyst at higher temperature and pressure, it was found that at the high residue conversion, the coke formation could be controlled even if the residue conversion is up to 90%. Then, a five-lump model by analytical method based on Monte Carlo technique was used to calculate the kinetic parameters, and the product distribution in the condition of high conversion was predicted successfully.     

中压加氢改质装置动态数学模型的开发

以加氢裂化集总动力学模型为基础建立了中压加氢装置的仿真培训系统,重点是表征反应器的开工过程,并开发相应的加氢精制和加氢裂化反应器动态数学模型,使其在较宽的工艺条件范围内与试验值吻合。在仿真培训系统的实际应用表明,所开发的模型逼真度高,可满足实时性的要求,能产生较好的培训效果     

基于半监督学习-多通道卷积神经网络的加氢裂化产品性质预测

提出了一种用于加氢裂化产品性质预测的半监督学习-多通道卷积神经网络(SSL-MCCNN),通过逐层卷积实现加氢裂化工艺流程空间域局部特征提取,并基于多通道采样实现了时域特征提取。在应对模型训练中由于产品性质数据量不足导致的小样本学习问题方面,基于教师-学生半监督学习(TS-SSL)生成虚拟样本集实现了数据扩充,进一步提升了模型预测性能。基于SSL-MCCNN对煤油-柴油加氢裂化工业装置重石脑油密度和柴油闪点预测的均方根误差(RMSE)分别为0.83和1.03,判定系数(R²)分别为0.90和0.98,与BP神经网络(BPNN)和径向基神经网络(RBFNN)相比,SSL-MCCNN在实现最小RMSE的同时达到了最优R²。实验结果表明,所提出的SSL-MCCNN有效提取了加氢裂化工艺流程的时空域特征,显著提升了模型预测性能。     

悬浮床加氢裂化集总动力学模型的研究

在不同反应温度、氢初压条件下,通过高压反应釜对克拉玛依常压渣油（KLAR）进行加氢裂化实验,以此模拟悬浮床加氢裂化过程,并根据实验数据及实际工艺中对各种轻油产品收率预测的需求建立了悬浮床加氢裂化六集总（气体、汽油、柴油、蜡油、减压渣油、焦）动力学模型,用matlab软件进行编程,采用最小二乘法对动力学参数进行估算,并进行误差分析。结果表明,建立的六集总动力学模型能很好的对各集总产品收率进行预测,计算结果与实验值基本吻合,大部分误差在5%以内。     

基于LSTM的润滑油加氢装置产品预测和基于RF的操作优化相关性分析

基于某炼油厂润滑油加氢装置的生产工艺及实际生产数据,采用Aspen HYSYS软件建立了润滑油加氢装置的生产过程机理模型,并用随机抽样的方法验证了模型的有效性;然后根据生产工艺条件运行机理模型,扩展了装置的产品预测数据集。对比BP神经网络,采用LSTM(Long Short-Term Memory)神经网络,以加氢裂化反应器入口温度、加氢异构反应器入口温度、加氢后精制反应器入口温度、加氢裂化后常压分馏塔塔顶温度、加氢异构常压分馏塔塔顶温度、加氢裂化反应器压力、加氢异构反应器压力为输入,以轻质润滑油常温常压下的质量流量、40 ℃运动黏度、闪点和倾点4个目标为输出建立了预测精度更高的产品预测模型。通过随机森林(RF, Random Forest)算法对该装置产品预测数据集的输入特征变量与输出目标变量进行了相关性分析,确定了特征重要度排序,得到了不同生产目标对应的优化操作方案。