FCC CLOSED-CYCLONE SYSTEM ELIMINATES POST-RISER CRACKING

Amos A. Avidan, Frederick J. Krambeck
Mobil Research & Development Corp.
Paulsboro, N.J.

Hartley Owen, Paul H. Schipper
Mobil Research & Development Corp.
Princeton, N.J.

A major improvement in high-temperature, short-contact time catalytic cracking has been commercialized by Mobil Corp.

The use of closed cyclones in fluid catalytic cracking (FCC) reactors has eliminated post-riser, nonselective thermal cracking.

The most dramatic effect has been to lower dry-gas components, such as methane and ethane, by more than 40%. Gasoline and distillate yields and gasoline octane are increased, while heavy fuel oil yield is decreased.

The closed-cyclone system was demonstrated in a large cold-flow model and installed in eight Mobil FCC units. Commercial experience helped fine tune operational guidelines.

Catalytic cracking technology has been evolving for more than 50 years, yet it has not become a mature technology.1 New challenges in residuum processing, tightening environmental regulations, and the need to improve profitability through advanced control and optimization are transforming FCC. Typical FCC yields are shown in Fig. 1. The evolution in yields of the most useful products (gasoline and light olefins) is shown in Fig. 2. The gradual improvement is the sum of small evolutionary changes in hardware, catalysts, and processes.

A true revolution occurred in the mid-1960s when Mobil scientists developed zeolite cracking catalysts which have in turn led to dramatic changes in FCC reactor design. The older silica-alumina amorphous catalysts required long residence times, while the new zeolite-containing catalysts benefitted from residence times as short as 1 sec.

FCC reactors were modified to complete the reaction in the reactor riser, and the reactor has become a stripping vessel and an enclosure to hold internal cyclones (Fig. 3).

Modern FCC reactor risers are very efficient, yet can be improved in several areas.

One major area of inefficiency is thermal cracking in the reactor vessel. Along with shorter contact times, cracking temperatures have been raised, and many FCC reactors now operate at more than 1,000 F.

As the temperature was raised, competition from nonselective thermal cracking in the reactor vessel became more significant. It then became imperative to redesign the product-catalyst separation system to eliminate thermal cracking.

THERMAL CRACKING

Thermal cracking does not progress through the same carbenium ion chemistry as catalytic cracking. Yields are different, and the most significant differences are illustrated in Fig. 4, which shows the impact of thermal cracking on actual FCC reaction products as measured in a laboratory reactor. As temperature is increased from 950 F. to 1,050 F., gasoline and distillate (G + D) yields decrease and heavy fuel oil (HFO) and light gas yields increase. The light gas contains considerable amounts of methane and ethane, which are not only low in value, but can also decrease FCC unit capacity because of gas plant limits.

The activation energy of thermal cracking is about four times higher than that of catalytic cracking (60 vs. less than 15 kcal/gmolK.), and so thermal cracking is much more sensitive to increasing temperature.

Post-riser thermal cracking contributes more than half the yields of such compounds as methane and ethane in today's short-contact time FCC riser-reactors.

REDUCED THERMAL CRACKING

Many designs to reduce nonselective post-riser cracking have been investigated, developed, and commercialized during the past 15 years. They all share the same basic concepts of quickly separating spent catalyst from cracked products, introducing the catalyst into a stripping zone, and minimizing the time the vapor products are held at high temperature before being quenched in the FCC products fractionator.

Early FCC riser separator designs tended to favor inertial rough-cut separation devices on top of the riser (such as a simple T), followed by reactor cyclones, as in the previous dense-bed reactor designs. (This is shown in the reactor design in Fig. 3.)

Many different types of rough-cut devices have been patented, and some of them have been commercialized. Some examples are in Fig. 5 The rough-cut separator is basically a low-efficiency cyclone that relies on momentum direction changes rather than fully developed centrifugal force as in a cyclone. The inertial force acting on solids particles is typically 10-30 times lower than in the cyclone, and separation efficiency is 80-98%, rather than more than 99.9% as in a cyclone. Another approach has been the use of riser cyclones instead of a rough-cut device. One interesting variation on the riser cyclone is the stripper cyclone, where stripping steam is introduced into the cyclone. One such design is illustrated in Fig. 6. Mobil experimented with this approach in the 1970s. But riser cyclones and rough-cut devices share the problem of having a large volume between the gas exit from the riser separation device and the entrance to the reactor cyclones. The residence time in this volume can be as long as a minute, and thermal cracking proceeds at essentially the riser top temperature, resulting in significant loss in G + D and an increase in gas make.

MOBIL'S CLOSED CYCLONES

Tracer tests on a commercial FCC reactor have shown that at least 40% of the products backmix extensively in the reactor vessel. The design Mobil's engineers developed to eliminate this backmixing is a closed-cyclone system.

In the closed system, the top of the FCC riser is directly connected with the cyclones to quickly separate the products from the catalyst and to remove the products into the fractionator.

Direct connection of the riser to the cyclones required meeting several design objectives:

A means to remove stripper gas (steam and stripped hydrocarbons) had to be provided.
New heatup and start-up procedures had to be developed because the new system had different thermal expansion properties.
A robust design and new operational guidelines to prevent carryover of solids to the fractionator had to be developed, and because the catalyst surge volume between the riser top and reactor outlet was cut dramatically, the margin of error available during upsets was reduced accordingly.
To verify adequately the operation of the system, the design had to be tested and demonstrated on a large scale.

The first design objective was met by careful design of alternate means to remove stripper gases. This is accomplished by an additional reactor cyclone, or properly designed vents.

The other mechanical design objectives were demonstrated in a large-scale, coldflow model of the closed-cyclone system at the Paulsboro research laboratory. This unique, 120-ft tall model is constructed partially in Plexiglas (Figs. 7 and 8), thereby permitting visual observation in addition to various experimental measurements.

The model simulates the FCC reactor riser, the reactor vessel with the closed-cyclone system, and the stripper bed. Solids are recirculated through a standpipe into the bottom of the riser.

Reactor heatup was evaluated by means of a computer simulation of the heat transfer processes during this phase of operation. The results provided design features and start-up procedures to ensure adequate temperatures before introducing feed to the riser.

Two other important elements of closed-cyclone technology are yield estimates and commercial experience. Yield estimates were important for gauging the potential advantages of the new technology and to effect appropriate changes in the product-recovery section.

The yield estimates were obtained via our FCC kinetic model, with a thermal cracking package developed from laboratory data. The required post-riser residence time information was obtained from the cold-flow experiments.

These were verified once the system was installed commercially by means of helium tracer tests (Fig. 9). The response to a pulse of tracer in an open system, without closed cyclones, shows an average residence time in the reactor vessel of nearly 40 sec, with a wide spread (large standard deviation), indicating considerable backmixing.

The same test performed after installation of closed cyclones shows a much narrower response.

COMMERCIAL EXPERIENCE

Closed-cyclone systems were installed in eight Mobil FCC units during the 1980s. Installation was easily accomplished during scheduled turnarounds.

Initial commercial experience showed:

The anticipated yield benefits were immediately achieved. Not all FCC hardware improvements provide clear benefits-a one to three percentage point increase in gasoline yield may be hard to detect. Yet the dramatic decrease in dry gas make is obvious, even upon reversibly engaging and disengaging the closed cyclone system (Fig. 10).
Heatup was rapid and trouble-free, as predicted by our heatup and start-up model.
Initially, some operational difficulties were experienced, with some carryover of solids. These were solved by applying commercial experience, together with hydrodynamic principles, in developing a detailed set of operational guidelines.

Adherence to those guidelines has eliminated the problems. Carryover of solids to the fractionator, as measured by bottoms ash content, has actually decreased in comparison with preclosed-cyclone performance.

YIELD BENEFITS

Estimated yield shifts, based on our kinetic model, and commercial audits from three refineries, show significant benefits (Table 1). Typically, G+D yield increases by 2.5 vol %, and C2 - (dry gas) yield decreases by 1 wt %. (This is a reduction of 40% in sulfur-free, dry-gas make.)

A refinery usually takes advantage of a reduction in gas make by increasing cracking severity. An example of the benefits of closed cyclones at constant dry-gas make is shown in Table 2.

Riser top temperature was increased by 30 F., increasing gasoline, distillate, and alkylate yields by 3 vol % and increasing research octane by 0.8 and motor octane by 0.6 numbers.

Another benefit of closed cyclones is the reduction in unwanted by-products. One such thermal by-product, butadiene, increases acid consumption in downstream alkylation.

The effect of closed cyclones on butadiene yields is shown in Fig. 11. The production of butadiene is reduced by over 50%.

Another advance in catalytic cracking technology commercialized by Mobil in the 1980s is the use of ZSM-5 octane-enhancing additive. ZSM-5 is a smaller-pore zeolite than the Y-zeolite used in cracking, and it promotes shape-selective cracking and isomerization of low-octane hydrocarbons. Dry-gas yield does not increase.

Because closed cyclones allow reduction in dry-gas make, the use of ZSM-5 can further enhance light olefins production within gas-handling limitations.

The installation of closed cyclones has improved yields by eliminating post-riser thermal cracking. It allowed Mobil FCC units to increase operating severities and to practice high-temperature, short-contact-time cracking. The cost of installing the closed-cyclone system is small, particularly when compared with the immediate benefits.

ACKNOWLEDGMENT

Many researchers and engineers at Mobil Research & Development Corp. have contributed to developing, testing, and successfully commercializing FCC closed cyclones. They include F. M. Buyan, M. Edwards, J. H. Haddad, B. M. Knickerbocker (who conducted the cold flow model tests), D. M. Nace, K. Schatz, and P. T. Sparrell (who developed the heatup model).

We are also grateful for the close cooperation and assistance of Mobil marketing and refining personnel.