FCC UNITS BENEFIT FROM RECENT CATALYST ADVANCEMENTS

Anne K. Rhodes
Refining/Petrochemical Editor

The fluid catalytic cracking (FCC) process has been the focus of much catalyst research in recent years. The resulting improvements in metals passivation and catalyst-pore accessibility have been confirmed in a number of trials at operating refineries around the world.

Highlights of these trials are:

A U.S. refiner found metals passivation effective at low metals loadings.
A Venezuelan refiner used a new metals-trap catalyst to maintain catalyst activity as equilibrium catalyst (Ecat) metals levels increased drastically.
Two refineries held successful trials of new FCC catalysts with greater pore accessibility.
A French refiner, with the help of its catalyst supplier, fine-tuned its FCC catalyst formulation to improve yield structure.

LOW METALS LOADING

Metals passivation was first used in a heavy oil cracking unit with about 3,800 ppm Ni and 6.900 ppm V, or a total metals (4Ni + V) level of 22,100 ppm. But over the years, passivation has been fine-tuned and used in FCC units with less and less metals on Ecat.

Recent findings indicate that metals passivation is effective even in feedstocks containing relatively low levels of metals contamination. An unpublished report by Phillips Petroleum Co. described the results of a trial in the FCC unit at a 190,000 b/d refinery. Table 1 shows the results of the runs.

Before passivation (Case 1), the Ecat contained a total metals concentration of 2,517 ppm (4Ni+V). Following the first 3 months of injecting an antimony-based passivator (Case 2), the Ecat contained 2,615 ppm total metals.

As hydrogen and coke yields decreased, the refiner increased FCC charge by 489 b/d to 29,224 b/d gas oil. And feed quality was reduced when heavy cycle oil from another FCC unit was mixed with the gas oil.

Hydrogen yield decreased 35%-from 53 scf/bbl to 34 scf/bbl-and dry gas yield declined by 5.6%. With metals passivation, the refiner was able to increase feed while remaining well within the coke-burning limit (Table 1).

Feed composition was gradually changed to a blend of gas oil and atmospheric resid from sweet crude. Even with 30% more metals on the catalyst (3,311 ppm total) and lower-quality feed, hydrogen yield remained about even with the base case. Because the feed now contained 13% resid, dry gas make increased by 11%.

Ecat micro activity test (MAT) levels remained steady during the test period of several months. And the refiner was able to keep coke, hydrogen, and dry gas production within the necessary limits while processing a wide range of feedstocks.

NEW METALS TRAP

Intevep S.A.'s new metals-trap catalyst was tested in the FCC unit at Corpoven S.A.'s 195,000 b/d refinery in Puerto La Cruz, Venezuela. Both Intevep and Corpoven are affiliates of Petroleos de Venezuela S.A.

Results of the study were presented at Akzo Nobel's catalyst symposium, held June 5-8, in Noordwijk, The Netherlands. The paper was written by Intevep's N.P. Martinez, D. Huskey, N. Gamarra, J. Lujano, and J. Velasquez.

During the trial, the Ecat metals level was increased to 7,000 ppm vanadium and 1,700 ppm nickel. Operating results include steady catalyst activity, decreased coke and gas make, and a 40% reduction in catalyst addition.

THE TEST

The test comprised three stages:

Stage 1-A fresh catalyst makeup rate of 4.0 metric tons/day (mt/d) of the new catalyst, called Intfcc-1, was used to replace about 40% of the unit's 130 mt catalyst inventory. Stage 1 was completed successfully and conversion increased by 2.3 vol %.
Stage 2-Ecat vanadium and nickel content was increased further, with a corresponding decrease in the catalyst addition rate from 4.0 to 3.0 mt/d.
Stage 3-Catalyst make-up rate was reduced to 2.5 mt/d and Ecat vanadium and nickel levels increased to, respectively, 7,000 ppm and 1,700 ppm, as shown in Fig. 1.

Throughput during the trial was 12,500 b/d until an increase in feedstock availability caused Corpoven to increase it to 14,000 b/d. The reaction temperature was maintained at 535 C.

During Stage 3, the feed vanadium level fluctuated between 6 and 10 ppm, nickel content was about 25 ppm, and sodium fluctuated between 2 and 6 ppm. A peak sodium level of 12 ppm occurred when the feedstock became contaminated with seawater.

Fig. 2 shows the yields of gasoline and alkylation feed, and the conversion, observed during Stages 2 and 3. It can be seen that, despite increasing metals levels on the catalyst, the unit was operating quite stably.

Although nickel increased to 1,700 ppm from 1,000 ppm, yields of dry gas and coke decreased slightly. Fig. 3 shows dry gas yield during the test.

Similarly, regenerator operating temperature declined from 710 C. at the beginning of the trial to 690 C. by the end.

ECONOMICS

Corpoven's economic evaluation considered two scenarios. The first scenario comprised the results obtained at the end of Stage 1, with 40% of the new catalyst in inventory. Benefits were estimated to be $1.4 million/year.

The second scenario involved the operating conditions obtained at the end of Stage 3. Although estimated benefits of this scenario were less-$800,000/year-two qualifications should be noted:

The reference case for this scenario was that obtained with the base catalyst at an addition rate of 4.5 mt/d and an Ecat vanadium level of 4,800 ppm.
No credit was given for processing a heavier, more contaminated feed during Stage 3.

PORE ACCESSIBILITY

P. van de Gender and K.Y. Yung of Akzo Nobel presented the performance results of new highly accessible catalyst at the Noordwijk symposium.

Improved accessibility to functional sites enables faster transport of hydrocarbons through the catalyst particles.

The functional sites in these catalysts are more effective, say the researchers, and the largest molecules benefit most from easier access to these acid sites.

Advantages of improved accessibility include: Better bottoms cracking, better resistance to poisons, improved strippability, and reduced overcracking and secondary reactions.

TEST A

A refiner who had been using Vision-409 MR containing ADZ-40 zeolite and a vanadium trap, switched to the improved ADZ-50 zeolite. Because better zeolite stability and improved accessibility enhance vanadium tolerance, the vanadium trap was omitted.

Table 2 shows the results of the catalyst switch. The feedstock was heavier during the test period, and feed density, refractive index, and sulfur content were greater.

The poorer-quality feed caused an increase in delta coke and regenerator temperature. But despite these factors, dry gas yield decreased.

Gasoline selectivity improved as well, as is illustrated by the 1.6% increase in gasoline yield. And, despite higher conversion, LPG yield did not increase.

This enabled the refiner to increase conversion without exceeding wet-gas compressor limits.

TEST B

This FCC unit was severely limited by regenerator temperature. The high regenerator temperatures also limited riser temperatures.

The reduction of delta coke was crucial for this refiner, as lower delta coke allows increased feed rate and riser temperature. And higher riser temperature increases gasoline octane.

The traditional method of reducing catalyst surface area without losing activity is to increase rare earth content. This allows a reduced zeolite level and surface area for the same activity, but reduces gasoline RON.

To solve these problems, Octavision-516 was chosen. Table 3 shows the results from this trial.

Ecat surface area decreased by 40 sq m/g without an increase in rare earth content or a loss of activity. Because the regenerator temperature was lower, the riser temperature and catalyst-to-oil ratio could be increased. These changes allowed a 3 wt % improvement in conversion, even at a greater feed rate.

RON and MON increased 0.8 octane numbers each and gasoline yield increased 1.2 wt %.

CATALYST SELECTION

M. Bourgogne and D. Chombart of France's Total Raffinage Distribution S.A. reported results from Total's Flanders refinery near Dunkirk, France, at the conference. A catalyst supplier helped the refiner fine-tune its FCC catalyst formulation to decrease delta coke.

Matrix activity is related to its surface area and acid-site density. Increasing matrix activity or content increases bottoms conversion, but also increases coke and gas make. Considering these factors, Total's supplier suggested three major areas on which to focus for improving coke selectivity:

Zeolite-to-matrix ratio
Matrix alumina or silica-alumina type, surface area, acidity, and pore-size distribution
Activity level.

In addition, the zeolite part of the catalyst had to be designed properly, meaning:

Zeolite crystals with few defects
Optimal nature, content, and distribution along the zeolite crystals of the extra-framework alumina species
Carefully tailored zeolite-framework silica-to-alumina ratio, rare-earth exchange, and unit cell size.

Additional desired catalyst characteristics include:

Special macropores (1,000 diameter) with low activity to crack large asphaltene molecules and allow the metals they contain to be deposited and passivated.
Mesopores (30-1,000 ) with higher active surface area for cleaving side chains from aromatics and naphthenic rings.
Small pores (

STEP 1

The first new catalyst Total chose was designed to optimize the zeolite-to-matrix ratio and use a matrix with improved coke selectivity, especially in the presence of high nickel levels. Total chose "Catalyst A," the pore-size distribution of which is displaced toward larger pores.

Operation on Catalyst A was characterized by an increase in the amount and heaviness of resid-the resid content of the feed varied between 20 and 60 wt %. But the regenerator temperature was maintained within unit constraints.

Catalyst consumption remained essentially constant (0.1 wt % on fresh feed) and metals increased from 3,950 to 5,710 ppm Ni+V, while antimony injection declined.

Conversion increased 2.9% and olefins yield increased. The dry gas yield remained constant, despite a 5 C. increase in reactor temperature.

These results led Total to a second trial, the aim of which was to process additional resid and fine-tune their LPG constraint at high reactor temperatures (especially when processing paraffinic resids).

STEP 2

LPG selectivity traditionally is reduced by decreasing the zeolite-to-matrix ratio, but this can deteriorate coke selectivity. Total's catalyst supplier advised holding the zeolite-to-matrix ratio constant and improving accessibility to acid sites. This approach was expected to improve coke selectivity and reduce overcracking of gasoline at constant conversion.

Operation on "Catalyst B" was characterized by an increase in heavy atmospheric resid feeds. Resid intake during Step 2 averaged 55 wt % and Conradson carbon increased from 1.4 to 2.5 wt %. But because of the heaviness of the feed, the charge rate declined.

Catalyst consumption increased from 0.12 wt % on fresh feed to 0.14 wt %, and the metals content increased to 6,580 ppm Ni+V. In addition, the V/Ni ratio increased from 0.6 to 1.3, and the catalyst contained no vanadium trap.

Unit conversion (221 C. + coke) and liquid conversion (C3-360 C.) decreased. Dry gas yield rose and, surprisingly, so did propylene yield. The authors say this may indicate that the thermal-to-catalytic cracking ratio increased.

STEP 3

The choice of a third catalyst was oriented toward the use of a matrix with improved coke selectivity. The accessibility to acid sites was maintained, but strippability needed to be improved.

During winter operation, Total's main fractionator experienced loading problems at high feed rates. Improving strippability can alleviate the load on the column by reducing the stripping steam rate, with only a minor deterioration in delta coke. The feed rate, therefore, could be increased.

In Step 3, the vanadium trap was reintroduced.

Table 4 shows feed properties, process conditions, catalyst data, and unit performance for the third trial. The outstanding points of operation on "Catalyst C," compared to base case, include:

Resid intake more than doubled, with Conradson carbon increasing from 1.1 to 2.2 wt % (Fig. 4).
Catalyst consumption increased from 0.1 wt % on fresh feed to 0.12 wt %.
Metals increased to 7,070 ppm Ni + V, at a V/Ni ratio of 0.6.
Sb/Ni ratio on Ecat decreased by a factor of 6 (Fig. 5).
Catalyst surface area decreased 20%.
Propylene and diesel production increased.
Bottoms conversion remained constant.

Total's Flanders refinery now can process 85% atm resid in its FCC fresh feed. And Conradson carbon levels up to 3.5 wt % can be handled at a high feed rate without exceeding regenerator temperature constraints.

These case histories show how careful catalyst selection can improve refinery operations and yield patterns. With the help of petroleum catalyst suppliers' innovative research and testing programs, refiners can improve their bottom lines without making big capital expenditures.