Apr-2015

Understanding FCC feeds and benefits to refinery operation

In this work, over one hundred FCC feed samples, originating from diverse geographic locations and consisting of assorted FCC refinery streams, were assembled.

Uriel Navarro, Michelle Ni and Dariusz Orlicki
Grace Catalyst Technologies

Article Summary

To assure applicability to industrial processes, the following refinery streams, which are frequently part of the riser feed blend, were included in this study: atmospheric and vacuum gas oil, light and heavy coker gas oil, hydrocracked residue, atmospheric residue, and deasphalted oil. All of these feeds were characterised for the following physicochemical properties: density, boiling point temperature distribution by SIMDIS, refractive index, conradson carbon, total and basic nitrogen, sulphur, and metal content. hydrogen content was assessed by H-NMR. Other specialised techniques, such as SARA, UV-vis, and mass spectrometry for selected samples characterisation were also employed to study the hydrocarbons types and the chemical composition. At the end, all feeds were evaluated in the ACE pilot plant using two different commercial FCC catalysts. These catalysts were laboratory-deactivated using CPS metal-free deactivation protocol. The rationale for using metal-free lab-deactivated catalysts, besides simplifying the analysis, was to focus on the interactions between catalyst active sites (zeolite and matrix) and hydrocarbon feed components during the cracking process.

Detailed statistical analysis of feed samples properties produced correlations that would allow FCC refinery engineers to infer other unknown feed properties when only limited feed characterisation information is available. Another aspect of this research resulted in the development of the tools for prediction of hydrocarbon type distribution in the feed simply based on just feed density and boiling temperature distribution data. However, the most useful finding is a set of correlations between different feed properties and products yield obtained from pilot plant experiments. Reliable predictions were obtained for different FCC products, dry gas, C3 and C4 olefins, dasoline and its quality expressed by RON and MON, LCO, bottoms, and delta coke. Because the pilot plant data is frequently used to evaluate the feed and catalyst interactions under standard conditions, the tools developed here would help refinery engineers understand the effect of each feed component on the heat balance, product yields, and gasoline quality. Knowing the feed quality and its theoretical cracking capabilities would allow FCC operators to make appropriate changes in operating conditions, such as catalyst addition rate and riser and feed preheat temperature, to maximise feed potential and useful product yields. Some practical examples were provided for applying such statistical models to aide field operations.

Introduction
FCC unit feed is the most important variable of the process, because it has the greatest impact on operating conditions, yield and product quality. A feed for any FCC unit basically consists of the following hydrocarbon families: paraffins and cycloparaffins, which are the saturates, aromatic hydrocarbons with a different number of aromatic rings, and resins and asphaltenes, in cases where residue is processed. Feed behaviour in the riser also depends on the mechanical and operational conditions of the unit, and the concentration and distribution of these hydrocarbon families.
Therefore, the feed quality depends on the following factors: quality and type of crude processed by the distillation units and the refining systems existing in the refinery. The products of these alternate processes, such as the thermal conversion units like delayed coking, flexicoking and visbreaking, the hydrogen adding units like gas oil and residue hydrotreatment, and the hydrocracking of gas oil, liquid-liquid deasphalting units like the DEMEX process, and streams from lubricant production are sent to the FCC unit.

One of the biggest advantages of the FCC process is the flexibility to process all kinds of streams that are normally sent as complex mixtures, where the mixing processes are not always efficient enough to ensure a completely homogeneous blend. In general terms, the FCC unit feeds worldwide are blends of the following streams: atmospheric and vacuum gas oils as the major components, atmospheric residues, coker and visbreaking gas oils, hydrocracking residues, hydrotreated gas oils and residues, furfural extracts, demetalised oil (DMO), etc. However, one of the greatest weaknesses and problems in refineries is the lack of physical-chemical analyses for proper feed characterisation and what is more important for engineers, supervisors, and operators, the ability to predict the impact of the feed on yield, operating conditions (heat balance) and product quality.

The purpose of this paper is precisely to provide the staff of FCC units with useful tools to quantify the potential yield of the feed determined by chemical composition, i.e. by the different hydrocarbon families. However, the most important thing is that these potential yields are calculated using properties typically analysed in any refinery laboratory in the world, such as specific gravity, distillation, sulphur content, refractive index, etc.

To conduct this study, GRACE selected more than 100 types of FCC feeds from all over the world with different compositions. These feeds contain all the streams mentioned earlier, which are normally used in the patterns to feed these units. °API was used as selection criterion, considering feeds between 11 °API and 32 °API.

Feeds were fully characterised by all the physical-chemical analyses mentioned herein. The types of hydrocarbons in the feeds were determined by SARA analysis, mass spectrometry, UV-Vis and 1H-NMR, to quantify hydrogen content, which is an important property of FCC feeds.

Potential feed yields were analysed in an ACE¹ unit under the following operating conditions: Reaction temperature 527°C (980°F), reaction time of 30 sec, feed flow 3 g/min, catalyst/oil ratio of 4.0, 6.0 and 8.0, adjusted by varying the amount of catalyst in the reactor. Two GRACE commercial catalysts were used, deactivated by the CPS unit at 793°C (1460°F)²; the properties of the deactivated catalysts are reported in3,4. It is important to mention that the catalysts were deactivated in the absence of metals, since the purpose of the study was to study the interaction of the active sites of the catalyst (matrix and zeolite) with the different types of hydrocarbons in the selected feeds, and the metal contaminants (Ni and V) interfere with the reactions of hydrogen transfer, dehydrogenation and catalytic activity, which was not in the interest of this research.

Results and discussion
Feed classification
In order to understand FCC feed composition, they have to be classified by hydrocarbon type. There are many methods to do so in literature^5,6,7. In this paper, we used Correlation Index (CI) and hydrogen content. Correlation Index⁶ is a property developed to classify crudes and petroleum fractions by the US Bureau of Mines, according to the following equation:

Correlation Index = 473.7*d + 48640/(K+273) - 456.8       Eq. 1

Where K is the mid-boiling point in °C and d is specific gravity. The meaning of this property is very similar to the K characterisation factor. However, unlike the K factor that ranges between 11.0 and 13.0, CI ranges from 0 to 100. In this wide range, n-heptane = 0, cyclohexane = 50, while benzene = 100, which leads to the conclusion that low values correspond to paraffinic feeds, values close to 50 are typical of naphthenic feeds and higher values are of feeds that have a larger proportion of aromatic hydrocarbons.