Naphtha pyrolysis, modelling

Shu, W. R., Ross, L. L., Pang, K. H., "A Naphtha Pyrolysis Model for Reactor Design11. Paper No 27d, the 85th National Meeting of AIChE, Philadelphia, June 1978. 

In a previous publication (12,13) the development of pyrolysis models for gas and liquid feeds was discussed. In a more recent paper, Shu and Ross (14) described a generalized model for predicting the rate of thermal decomposition of naphthas. These models become key components of Braun s computational system. 

Ross and Shu [38] report a model for naphtha pyrolysis. Naphtha feed is treated as a single lumped constituent A, which decomposes by a pseudo-elementary process of the first order into products By, viz. 

Computer modeling of hydrocarbon pyrolysis is discussed with respect to industrial applications. Pyrolysis models are classified into four groups mechanistic, stoichiometric, semi-kinetic, and empirical. Selection of modeling schemes to meet minimum development cost must be consistent with constraints imposed by factors such as data quality, kinetic knowledge, and time limitations. Stoichiometric and semi-kinetic modelings are further illustrated by two examples, one for light hydrocarbon feedstocks and the other for naphthas. The applicability of these modeling schemes to olefins production is evidenced by successful prediction of commercial plant data. 

Semikinetic modeling is illustrated by a generalized model for naphtha pyrolysis. The empiricism associated with the semikinetic model dictates the need for an extensive data base for parameter estimates. The naphtha data base consists of about 400 tests covering pure components and their mixtures and 17 naphthas (25). The pure components studied were normal and isopentanes, cyclohexane, and n-heptane. The wide range of naphtha feed properties is summarized in Table III. 

Feed Decomposition. In the case of naphtha pyrolysis, feed conversion is not generally known. Direct measurement is precluded by the large number of feed components and the difficulty of analyzing for them in the reactor effluent. To overcome this problem, a model was developed to relate the C3 and lighter yield, an experimentally accessible variable, to feed conversion. 

Successful stoichiometric modeling is demonstrated for industrial pyrolysis of light hydrocarbon and their mixtures. A semikinetic approach is more appropriate for naphtha pyrolysis. Although the final form of such a model is simple, its development generally requires more innovations. Applicability of the naphtha model to olefins production is evidenced by the successful prediction of commercial plant performances. 

Based on naphtha pyrolysis experiments conducted in a bench scale tubular reactor (suitable for the simulation of industrial tubular furnace operations and taking into account the changes of expansion, temperature and the pressure in the reactor), a kinetic model has been developed for the calculation of the degree of de-con osition, the actual residence time, and the severHy of cracking. 

The first was presented by Kumar and Kunzru (1985) for modelling of naphtha pyrolysis. In this study, it is assumed that naphtha could be represented as a pseudo-pure compound and the primary decomposition represented by a single reaction with the initial selectivities determined experimentally. Based on the experimental results, the primary reaction is represented by a first order reaction for the whole range of conversions and the initial selectivities are assumed to be constant. The secondary reactions are also represented by molecular reactions and only important secondary reactions that can occur between the various primary products have been accounted for. A sixth-order Runge-Kutta-Verner method was used by these authors to solve a set of reactions which include 22 reactions and 14 components. The predicted and experimental product yields were compared and the rate constants adjusted, by trial-and-error, to minimize the deviation between the predicted and experimental values. The major limitation of this approach is that the initial selectivities are to be determined experimentally. 

Mathematical models for the pyrolysis of naphthas, gas oils, etc. are relatively empirical. The detailed analysis of such a feedstock is essentially impossible, and all heavier feedstocks have a wide range of compositions. Such heavy hydrocarbons also contain a variety of atoms often including sulfur, nitrogen, oxygen, and even various metal atoms. Nevertheless, certain models predict the kinetics of pyrolysis, conversions, yields, etc. with reasonable accuracy and help interpret mechanistic features. 

Naphtha feed is treated as a single pseudo species. Naphthas, used as pyrolysis feedstocks, are mainly composed of paraffins and naphthenes, with lesser amounts of aromatics. Olefin content is usually very small. Consistent with observed pyrolytic behavior of paraffins and naphthenes (15,16,26,27,28), feed decomposition is assumed to follow first-order kinetics. Equation 3 of the reactor model can be simplified as follows. 

At this point, the yield model can be used to represent the pyrolysis behavior of specific feedstocks. To generalize the model, the effect of feed properties must be incorporated. Naphthas are complex mixtures of hydrocarbons. Feed characterization is needed to condense a detailed naphtha description into a manageable set of parameters, which uniquely defines feed-dependent conversion and yield effects. 

The use of computer generation systems in modelling the pyrolysis of large hydrocarbons is no longer considered simply an alternative to manual mechanism construction. It has become a necessity. The quantity of species and reactions becomes enormous, increasing molecular weight. This is particularly true if the focus is not merely on linear alkanes but also on other typical components of naphthas and gasoils, such as Bo-alkancs or cyc/o-alkanes, where the number of possible isomers increases exponentially with the number of carbon atoms in the molecule. 

Detailed kinetic schemes also consist of several hundreds of species involved in thousands of reactions. Once efficient tools for handling the correspondingly large numerical systems are available, the extension of existing kinetic models to handle heavier and new species becomes quite a viable task. The definition of the core mechanism always remains the most difficult and fundamental step. Thus, the interactions of small unsaturated species with stable radicals are critical for the proper characterization of conversion and selectivity in pyrolysis processes. Parallel to this, the classification of the different primary reactions involved in the scheme, the definition of their intrinsic kinetic parameters, the automatic generation of the detailed primary reactions and the proper simplification rules are the important steps in the successive extension of the core mechanism. These assumptions are more relevant when the interest lies in the pyrolysis of hydrocarbon mixtures, such as naphtha, gasoil and heavy residue, where a huge number of isomers are involved as reactant, intermediate and final products. Proper rules for feedstock characterizations are then required for a detailed kinetic analysis. 

A kinetic-mathematical model is presented here that was developed based on experimental pyrolysis data obtained on straight run Romashkino (Soviet Union) crude naphtha cuts. Characteristics of this naphtha feedstock are summarized in Table 1. 

---