When studying the evolution of refractory species during hydrotreatment, the polyheteroatomic compounds containing both nitrogen and sulfur or nitrogen and oxygen atoms are sometimes neglected. However, even a low amount of these compounds could reduce the efficiency of the removal of N1 or S1 compounds, which is the main target of hydrotreatment. In this paper, three different vacuum gas oil feedstocks were analyzed along with their corresponding hydrotreated samples using electrospray ionization (ESI(+) and ESI(−)) Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) analyses. The evolution of the aromaticity and the alkylation of the polyheteroatomic compounds during hydrotreatment were followed using double bond equivalent (DBE) versus the number of carbon atoms (DBE = f(#C) plots). Moreover, as the hydrotreated samples were produced using the same operating conditions, the reactivity of the different feeds has been assessed. Finally, an additional hydrotreated sample was produced using a higher temperature and analyzed, allowing us to monitor the impact of temperature on the removal of the heteroatomic species. From a general point of view, basic polyheteroatomic species are more refractory than neutral or acidic polyheteroatomic species. It has also been shown that depending on the heteroatomic class considered, the hydrotreated samples could exhibit different aromaticity and alkylation distributions, whereas the distributions of the feeds were similar, which is characteristic of different hydrotreatment mechanisms. Finally, the positive impact of higher operating temperature on the removal of most of the species but in different ratios has been clearly demonstrated. This study takes advantage of FT-ICR MS to monitor the evolution of refractory species during hydrotreatment as well as being a powerful tool to put forward the reactivity differences between several feedstocks or study the impact of hydrotreatment operating conditions.
This work concerns the development of a methodology for kinetic modelling of refining processes, and more specifically for vacuum residue conversion. The proposed approach allows to overcome the lack of molecular detail of the petroleum fractions and to simulate the transformation of the feedstock molecules into effluent molecules by means of a two-step procedure. In the first step, a synthetic mixture of molecules representing the feedstock for the process is generated via a molecular reconstruction method, termed SR-REM molecular reconstruction. In the second step, a kinetic Monte-Carlo method (kMC) is used to simulate the conversion reactions on this mixture of molecules. The molecular reconstruction was applied to several petroleum residues and is illustrated for an Athabasca (Canada) vacuum residue. The kinetic Monte-Carlo method is then described in detail. In order to validate this stochastic approach, a lumped deterministic model for vacuum residue conversion was simulated using Gillespie's Stochastic Simulation Algorithm. Despite the fact that both approaches are based on very different hypotheses, the stochastic simulation algorithm simulates the conversion reactions with the same accuracy as the deterministic approach. The full-scale stochastic simulation approach using molecular-level reaction pathways provides high amounts of detail on the effluent composition and is briefly illustrated for Athabasca VR hydrocracking.
Abstract Ultra high-resolution mass spectrometry (FT-ICR MS) coupled to electrospray ionization (ESI) provides unprecedented molecular characterization of complex matrices such as petroleum products. However, ESI faces major ionization competition phenomena that prevent the absolute quantification of the compounds of interest. On the other hand, comprehensive two-dimensional gas chromatography (GC × GC) coupled to specific detectors (HRMS or NCD) is able to quantify the main families identified in these complex matrices. In this paper, this innovative dual approach has been used to evaluate the ionization response of nitrogen compounds in gas oils as a case study. To this extent, a large gas oil dataset has been analyzed by GC × GC/HRMS, GC × GC-NCD and ESI(+/−)-FT-ICR MS. Then, the concentrations obtained from GC × GC-NCD have been compared to those obtained from FT-ICR MS hence proving that strong ionization competitions are taking place and also depending on the origin of the sample. Finally, multilinear regressions (MLR) have been used to quantitatively predict nitrogen families from FT-ICR MS measurements as well as start rationalizing the ionization competition phenomena taking place between them in different types of gas oils.
Advanced characterization of the products of the hydrotreatment of gas oils is of high interest for refiners and can be achieved using ultrahigh resolution mass spectrometry (FT-ICR MS). However, the analysis of gas oil samples by FT-ICR MS generates complex data sets with numerous variables whose exhaustive analysis requires the use of multivariate methods. Relevant information about nitrogen and sulfur compounds contained in several industrial gas oils are obtained by using three different ionization modes that are electrospray ionization (ESI) used in positive and negative polarities and atmospheric pressure photoionization (APPI) used in positive polarity. For data sets generated for a single ionization mode, classical multivariate methods such as Principal Component Analysis (PCA) are commonly used. When the key information is spread into several ionization modes and thus into several data sets, a data fusion approach is highly interesting to simultaneously explore these data sets and can be followed by Parallel Factor analysis (PARAFAC). Nevertheless, many more variables are simultaneously considered when data fusion is performed and the sensitivity of PARAFAC and its ability to extract the most relevant variables compared to classical multivariate methods has not been assessed yet in the framework of FT-ICR MS. In this paper, a comparison of the classical data analysis (PCA) approach and the data fusion combined with the PARAFAC analysis approach is presented. The results have shown that applying PARAFAC on fused data sets is highly sensitive and able to put forward features and variables that are individually identified through classical data analysis with greater ease of implementation and interpretation of results. As an example, dibenzothiophenes and carbazole families (DBE 9) have explained most of the variance between samples and remain the most refractory compounds in hydrotreated samples. A significant difference in alkylation between the different types of gas oils has also been spotted. This paper validates the power and efficiency of this approach to explore complex data sets simultaneously without any loss of significant information.
In this paper, kinetic modeling techniques for complex chemical processes are reviewed. After a brief historical overview of chemical kinetics, an overview is given of the theoretical background of kinetic modeling of elementary steps and of multistep reactions. Classic lumping techniques are introduced and analyzed. Two examples of lumped kinetic models (atmospheric gasoil hydrotreating and residue hydroprocessing) developed at IFP Energies nouvelles (IFPEN) are presented. The largest part of this review describes advanced kinetic modeling strategies, in which the molecular detail is retained, i.e. the reactions are represented between molecules or even subdivided into elementary steps. To be able to retain this molecular level throughout the kinetic model and the reactor simulations, several hurdles have to be cleared first: (i) the feedstock needs to be described in terms of molecules, (ii) large reaction networks need to be automatically generated, and (iii) a large number of rate equations with their rate parameters need to be derived. For these three obstacles, molecular reconstruction techniques, deterministic or stochastic network generation programs, and single-event micro-kinetics and/or linear free energy relationships have been applied at IFPEN, as illustrated by several examples of kinetic models for industrial refining processes.
A vacuum residue is a complex hydrocarbon mixture of several thousand different chemical species. Even today, no analytical technique is powerful enough to obtain the molecular detail that is required for the development of a detailed kinetic model. To overcome this drawback, a two-step reconstruction algorithm has been developed to build a representative set of molecules from partial analytical data. The first step, called Stochastic Reconstruction (SR), creates an initial mixture of molecules by a Monte Carlo sampling method. The second step, termed Reconstruction by Entropy Maximization (REM), modifies the molar fractions of the molecules in order to improve the representativeness of the generated mixture. The combined SR-REM algorithm creates a synthetic blend of molecules whose mixture properties are close to the analytical data of the petroleum fraction. The method has been applied to petroleum vacuum residue fractions from four different geographic locations with substantially different compositions. All cases are well represented, clearly illustrating the versatility of the SR-REM method. As an extension to this base algorithm, a novel indirect two-step reconstruction algorithm was developed, in which the SR step is used to build a single reference mixture. The set of molecules thus obtained is subsequently used in the second step to represent various petroleum fractions via the REM method. This allows to simultaneously reduce the computational burden and to represent the vacuum residue fractions with the same set of molecules. To validate this alternative approach, eight vacuum residues from different origins have been reconstructed with this technique. The results in terms of analysis prediction have shown a very good agreement.
