Prediction feedstock composition

However, it has become apparent, with the introduction of the heavier feedstocks into refinery operations, that these ratios are not the only requirement for predicting feedstock character before refining. The use of more complex feedstocks (in terms of chemical composition) has added a new dimension to refining operations. Thus, although atomic ratios, as determined by elemental analyses, may be used on a comparative basis between feedstocks, there is now no guarantee that a particular feedstock will behave as predicted from these data. Product slates cannot be predicted accurately, if at all, from these ratios. 

Fractionation of heavy feedstocks into components of interest and study of the components appears to be a better approach than obtaining data on whole residua. By careful selection of a characterization scheme it may be possible to obtain a detailed overview of feedstock composition that can be used for process predictions. 

The stated goal was to develop RCC product yield predictions from feedstock composition and properties. One of the first relationships investigated was a coke yield relationship. Table IV shows the relationship first observed between coke production and feedstock composition. As can be seen, for this particular feedstock the average coke make over a total of 24 tests was 13.5 wt, % while the amount of polar aromatics plus asphaltenes in the feedstock was 12.5 wt, %. This is a fairly close approximation of coke yield. 

Under the processing agreement with the Crown, NZSFC is required to predict the operating efficiency of the complex for varying feedstock compositions. This is achieved using the Aspen Process Simulator which has been jointly developed by NZSFC and Davy McKee for this purpose. 

The properties of the zeolite play a significant role in the overall performance of the catalyst. Understanding these properties increases our ability to predict catalyst response to changes in unit operation. From its inception in the catalyst plant, the zeolite must retain its catalytic properties under the hostile conditions of the FCC operation. The reaclor/regenerator environment can cause significant changes in chemical and structural composition of the zeolite. In the regenerator, for instance, the zeolite is subjected to thermal and hydrothermal treatments. In the reactor, it is exposed to feedstock contaminants such as vanadium and sodium. 

Therefore, in this chapter we describe major refinery operations and the products therefrom and focus on their composition, properties, and uses. This presents to the reader the essence of petroleum processes, the types of feedstocks employed, and the product produced, as well as warning of the types of the chemicals that can be released to the environment when an accident occurs. Being forewarned offers an environmental analyst the ability to design the necessary test methods to examine the chemical(s) released. It offers environmental scientists and engineers the ability to start forming opinions and predictions about the nature of the chemical(s) released, the potential effects of the chemical(s) on the environment, and the possible methods of cleanup. 

Thus, fractionation methods also play a role, along with the physical testing methods, of evaluating heavy oils and residua as refinery feedstocks. For example, by careful selection of an appropriate technique it is possible to obtain a detailed map of feedstock or product composition that can be used for process predictions (Chapter 3). 

Predicting processabilty is a matter of correlating the test data (Chapter 2) with an understanding of the chemical and physical composition of the feedstock as it relates to properties, rehning behavior, and product yields (Speight, 1992). The molecular complexity of the heavy feedstocks offers disadvantages to the reliance on the use of bulk properties as the sole means of predicting behavior (Dolbear et al. 1987). 

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. 

Carbon deposition is one of the luost serious problems of the steam reforming catalyst process (ref 1). The deposition of carbon on naphtha steam reforming catalysts depends ori the chemical composition of the hydrocarbon oil, the steam/carbon ratio in the feedstock, as well as the pi ocesa temperature and pressure, it is also affected by tlie presence of sulfur poisons Our past research of SNG catalysts ejiamined the nature of the carbon deposits as a function of the sulfur level on the catalyst (refs, 2 4). A small amount of sulfur was found to promote the formation of carbon that is non-reactive with steam and hydrogen under steam reforming reaction conditions. The continuous accumulation of this less reactive carbon [continuous carbon deposition (CCD)l on the catalyst surface leads to coke fouling Studies of the occurrence of CCD in our laboratory tests allow ua to predict, that coke fouling is likely to occur on the same catalyst used in real Indusl.rlal applications. 

The influence of geometrical factors, the specific reactorload and the feedstock properties on the fuel gas production including the amount of tarry components will be analysed. No atttention will be paid to the possibility of in situ sulphur or chlorine removal. Earlier attempts to predict the product gas composition were based on thermodynamic models i.e. a combination of mass and energy balances assuming for one or more reactions chemical equilibrium at an empirically determined temperature e.g. outlet temperature. 

We are tempted to proceed a little bit further, and examine the development of the whole flowsheet in relation with the reaction system. Let s suppose that the feedstock is of high purity ethylene and benzene. Because recycling a gas is much more costly than a liquid, we consider as design decision the total conversion of ethylene. The benzene will be in excess in order to ensure higher conversion rate, but also to shift the equilibrium. The equilibrium calculation can predict with reasonable accuracy the composition of the product mixture for given reaction conditions. Then polyalkylates, mainly diethylbenzene can be reconverted to ethylbenzene in a second reactor. 

Additional studies in progress include a survey of water gas shift and cracking catalysts, one pass steam effect (no pyrolysis recycle) and a survey of various feedstocks. Commercial water gas shift catalysts are limited to a maximum temperature of about 900°F and thus are not appropriate for fluidization at the temperatures under investigation. A fixed bed in the overhead system however with steam feed significantly altered the product composition (Table IV) as predicted ... 

History. Since their introduction of catalytic reforming in the 1940s, models have been developed to predict the response of the units to process and feedstock changes. The evolution of the models have followed the increasing ability of refiners to analyze feed and product compositions and the use of computers to rapidly produce answers from increasingly complex calculations. 

Traditional hydrocarbon conversion process models have implemented lumped kinetics schemes, where the molecules are aggregated into lumps defined by global properties, such as boiling point or solubility. Molecular information is obscured due to the multi-component nature of each lump. However, increasing environmental concerns and the desire for better control and manipulation of the process chemistry have focused attention on the molecular composition of both the feedstocks and their refined products. Modeling approaches that account for the molecular fundamentals underlying reaction of complex feeds and the subsequent prediction of molecular properties require an unprecedented level of molecular detail. 

---