Bottom-cracking additive improves FCC middle distilate yeilds

For the first time, bottom-cracking additives (BCA) have been tested in a distillate-mode fluid catalytic cracking (FCC) unit.

The research and development center of Indian Oil Corp. Ltd. conducted the BCA trial in one of its distillate-mode FCC units. Plant data demonstrate that BCA can improve the middle distillate yield by about 2 wt % or more (depending on feed heaviness) through selective upgrading of the bottoms.

Bottom-cracking additive is a large pore (50 Å), noncrystalline, silica-alumina matrix with sufficient acidity to precrack larger molecules. This additive is a very recent development in the FCC field, and only a few commercial trials have been done so far, most in gasoline-mode FCC units.

FCC is the major secondary conversion process in Indian refineries. Conversion of bottom fractions and maximization of middle distillate yield are two major objectives of Indian FCC units. Because middle distillate is an intermediate product in the overall cracking process, per-pass conversion in an FCC unit is to be maintained at about 50 wt % to maximize the yield of middle distillate. This limitation in per-pass conversion, however, leads to a higher yield of unconverted bottoms, i.e., above 10 wt %.

Indian FCC unit operations

Distillate-mode FCC operations lead to higher amounts of unconverted bottoms than that of gasoline-mode FCC units.

Indian distillate-mode FCC units are generally equipped with a partial combustion regenerator. The units are usually operated at low reaction temperatures (490-500° C.) and have low catalyst-to-oil ratios (4.5-6) to keep per-pass conversion around 45-50 wt %. Low regenerator temperature, i.e., 630-660° C., and restricted air supply to the regenerator lead to higher coke on regenerated catalyst (CRC) of 0.30-0.60 wt %.

High CRC and moderately low riser top temperatures restrict the overcracking of the middle distillate produced. However, this restriction puts limits on the unit conversion level of about 45-50 wt % per-pass for middle-distillate yield.

The process results in a considerable amount of unconverted bottoms, part of which is recycled back to the riser reactor for further cracking. Recycling is done within the limits of allowable regenerator temperature and the air blower capacity, but the unconverted-bottoms yield is still about 5-10 wt % more than that of gasoline-mode operation.

In gasoline-mode operation, the per-pass conversion is very high (70 wt %). The reaction temperature is 520-540° C., and the regenerator temperature is 700-750° C. Catalyst-to-oil ratios are also high, 8-12. The regenerator is operated at full-combustion mode so that CRC is close to zero.

CLO reduction in distillate-mode

Clarified oil (CLO) is the unconverted product in the FCC unit. In distillate-mode, CLO can be reduced by increasing the operating severity of the unit, selection of proper catalyst, or using an additive.

Clarified oil (CLO) yield of a distillate-mode FCC unit may be reduced simply by increasing reaction severity to be like that of gasoline mode. This method is not advisable because distillate-mode FCC units are not designed for high-severity operations. Second, high-severity operations will shift the yield towards gasoline maximization rather than distillate maximization, which is the overall objective.

Reduction in CLO can also be done with proper catalyst selection. An increase of zeolite content in the host catalyst is not desirable. The kinetic diameters of the heavier molecules in vacuum gas oils are so big that they cannot enter the small zeolite pores. As a result, these heavier molecules are nonselectively (i.e, thermally) cracked to undesirable coke and dry gas.

The addition of a nonzeolite active catalyst in the host catalyst improves the conversion of heavier fractions. The active-matrix pores are large enough to allow the entry of the heavier molecules for selective catalytic cracking. This is a feasible option for refineries that process feedstock with constant quality.

Because feed quality depends on feed source, however, the availability of constant feed quality processed in a particular FCC unit cannot be ensured. More than 50% of feed being processed in Indian Oil refineries is imported.

Change in feed quality requires change of the base catalyst formulation to form a suitable zeolite-to-matrix ratio, which means replacement of the entire catalyst inventory. Switching over to a new catalyst formulation is not always possible to improve bottoms conversion.

Separate solid particle approach

Additive use in an FCC unit to meet various seasonal demands ^1-3 is common. For example, ZSM-5 additive can help FCC units meet sudden increases of LPG demand. Similarly, other additives can counteract the effect of fluctuations of feed properties , e.g., higher nickel and vanadium levels which poison the FCC catalyst.

In a similar way, Mitchell, et. al.,¹ adopted a separate solid particle additive approach to upgrade the bottom of the barrel by adding bottom-cracking additive with host catalyst. He developed this option from laboratory studies. Since Mitchell's work, bottom-cracking additives have been successfully used in several commercial units running in high severity gasoline mode.^4-5

Bottom-cracking additive

As mentioned, conventional zeolites cannot convert feed molecules effectively. Feed molecules boiling greater than 485° C. are prevented from entering into the Y-zeolite cage because the kinetic diameters of these molecules are much bigger than the zeolite-pore opening. ¹

Adding nonzeolite active catalyst to the host catalyst can improve the selective cracking of heavy molecules of feed components. The heavy molecules are first cracked on the surface of the catalyst matrix. The products of this cracking, which are small, are further cracked by zeolites to produce lighter products.⁶

During the primary cracking on matrix surface, contaminant metals, such as nickel and vanadium, are deposited on the surface. Therefore, any further reactions on the matrix surface encourage dehydrogenation and poly condensation reactions, resulting in more dry gas or more CLO yield.

The bottom-cracking additive is a matrix based on very high alumina with highly acidic activity. It provides alternate active sites in which heavy feed molecules are allowed to enter and crack in a selective manner into smaller fractions. These smaller fractions are then upgraded by Y-zeolites present in the host catalyst.

Because both the BCA precracking of heavier molecules and the subsequent upgrading by Y-zeolites are via catalytic cracking routes, a reduction in coke and dry gas formation occurs.

Incidentally, BCA is expected to reduce the SO_x emissions in FCC-regenerator flue gas. A large percentage of the active component used to prepare BCA is gamma alumina, which has SO₃ capturing ability. This SO_x emission reduction ability is an additional advantage to those FCC units in which reduction of SO_x emission is of concern.

BCA experiment

In this experiment, ReY-type catalyst was used as host catalyst. Both the additive and catalyst were separately steamed at 788° C. for 3 hr in a 100% steam atmosphere. Performance of the BCA was evaluated at different concentrations (0, 5, 7.5, 10, and 20 wt %) with the host catalyst in a laboratory fixed-bed micro-reactor, which simulated a distillate-mode FCC operation.

Simulations of distillate mode FCC units in a laboratory micro-reactor are well-established.³ FCC catalysts and additives of most Indian FCC units are being recommended based on simulated micro-reactor data.

The pretreated catalyst and additive were mixed together at different concentrations for the evaluation. The product selectivities with BCA additive were established from the simulated data taken at different reaction severity levels in the micro-reactor. The feed used for this study was the actual combined feed being processed in one commercial DCC unit. The feed properties are summarized in Table 1 [11,237 bytes].

The results obtained from a simulated micro-activity test (MAT) unit were fed to an in-house developed FCC simulator (Fccmod) for the performance prediction of the additive at equal coke levels like those observed in the heat balanced commercial FCC unit. The capability of Fccmod for catalyst and additives performance prediction is well-documented.³

Based on the performance prediction of the BCA at different concentrations, a plant trial was recommended. Laboratory and plant trial results are discussed in subsequent sections.

Laboratory results

In Fig. 1 [25,789 bytes], the yield of total cycle oil (TCO), coke, and bottoms (i.e, materials having a boiling point greater than 370° C.) are plotted against different concentrations of the BCA at a conversion of 42 wt %. TCO is the stream which goes to diesel, commonly called middle distillate. The conversion is defined as the weight percentage of products boiling less than 216° C., including coke.

It is interesting to note that TCO yield increases with increasing BCA concentration in catalyst inventory, up to about 5 wt % BCA. Simultaneously, coke and bottoms yields reduce to a minimum at about 5 wt % BCA. After attaining the maximum TCO yield with minimum bottoms and coke yields, there is sharp reduction of TCO with a simultaneous increase in coke and bottoms yield.

The increment of coke yield is very sharp compared to the TCO reduction or the bottoms increment. This phenomenon establishes an optimum level of BCA (about 5 wt %) at which the bottoms and coke yield are a minimum and where TCO is a maximum.

The increase of TCO along with reduction of coke and bottoms yield is due to the reduction of thermal cracking as the BCA concentration increases from 0 to 5 wt %. In the absence of BCA or in the presence of a lower concentration of BCA, the bigger molecules get thermally cracked to mostly coke, dry gas, and CLO because they are prevented from entering into the zeolite pore for selective cracking.

With the increase in BCA concentration, selective cracking of bigger molecules is enhanced and reaches an optimum concentration (about 5 wt %) when the zeolite-to-matrix ratio encourages maximum catalytic cracking.

With a further increase of BCA concentration, thermal cracking is predominant, which leads to increased coke and bottoms yield. The optimum concentration of BCA is dictated by the feed heaviness, the host catalyst composition, and the operating severity level.

Based on simulated MAT results and physical and chemical properties of the BCA (e.g., the alumina content), in-house Fccmod predicted the commercial plant performance of the BCA. Subsequently, a plant trial was conducted. The predicted performance of this additive is summarized in Table 2 [19,374 bytes].

Fccmod predicts that with BCAs, heavy naphtha (150-216° C.) is likely to increase by 0.67 wt %, light cycle oil (216-370° C.) is likely to increase by 1.48 wt %, and CLO is likely to decrease by 2.27 wt %.

At the same time, the regenerator dense-bed temperature will likely decrease by about 11° C. due to lower coke make (Fig. 1).

Due to the low-severity operation of a distillate-mode FCC unit, the bottoms are upgraded to a TCO product and do not further crack to gasoline and LPG. In contrast, in a gasoline-mode FCC unit, the bottoms are converted to a TCO range product, which is further cracked to a lighter product like gasoline.⁵

Due to the high alumina content of BCA, SO_x emission in flue gas is expected to be reduced by about 150 ppm, or 30 vol %.

Plant trial

The BCA plant trial was conducted in a UOP-licensed, side-by-side, riser-type FCC unit. The unit has a feed-processing capacity of 1 million tons/year and a catalyst inventory of 80 tons. It is operated in distillate-maximization mode.

First, Indian Oil wanted to establish base-case properties of feed, catalyst, and additive without the use of BCA. Table 3 [9,465 bytes] summarizes the properties of the feed processed during the plant trial, which was a mixture of imported heavy vacuum gas oil (HVGO) and Bombay High HVGO.

The base catalyst was an ReY type. Fresh catalyst addition rate was about 0.74 tons/day during the trial period. An LPG boosting additive was used throughout the plant trial run at the rate of 15 kg/day. Contaminant metal content of nickel and vanadium on equilibrium catalyst was about 1,000 and 2,100 ppm, respectively.

From Dec. 3, 1994, to Dec. 30, 1994, BCA was added for the commercial plant trial. The BCA dosing rate started at a rate of 300 kg/day until achieving a 3 wt % BCA concentration in the inventory. Thereafter, dosing was maintained at the rate of 30 kg/day to maintain 3 wt % BCA concentration in catalyst inventory. Although the recommended BCA concentration was 5 wt %, the trial run was conducted with 3 wt % BCA due to insufficient BCA inventory.

After the BCA concentration had been maintained at 3 wt % for about 1 week without varying other process conditions, a test run was conducted. The yields and process conditions of the base-case vs. the test run with 3 wt % BCA are compared in Table 4 [11,486 bytes].

Table 4 shows that TCO yield is increased by about 1.71 wt % with the reduction of 1.66 wt % CLO. There is no remarkable change in other product yields. A comparison of Indian Oil's prediction (Table 2) and plant trial results reveal that the incremental shifts of TCO and CLO in the plant trial are relatively less than expected. This difference results from a lower concentration of BCA in catalyst inventory at the trial condition, 3 wt % rather than 5 wt %.

In this trial, SO_x emission reduction is only 15 vol % of the base BCA performance. The expected reduction was about 30 vol %. Again, this lower reduction of SO_x emission is due to the lower concentration of BCA in inventory.

Because coke make is less, some decrease in regenerator temperature was expected. The plot of regenerator dense-bed temperature before and after the trial run is plotted in Fig. 2 [19,374 bytes]. From this figure, it can be observed that there is a drop of regenerator dense-bed temperature by about 4-8° C., as predicted from laboratory data. This reduction of temperature can be used to increase unit throughput.

Similarly, product properties of the base case and the trial have been compared. Large changes in diesel cetane number, CLO density, and sulfur in gasoline and and diesel have not been observed. These observation are in line with BCA trials reported in literature.^{4 5}

Distillate vs. gasoline modes

The incremental differences in yields and operating conditions for distillate-mode vs. gasoline-mode FCC units are shown in Table 5 [20,740 bytes]. The product-yield types and measurement units for distillate and gasoline modes are different. The units for all product yields for gasoline mode are vol % except fuel gas and coke, which are expressed in wt % of fresh feed.

The feed properties used in the distillate-mode FCC during the test run are listed in Table 3. Gravity of feed used in gasoline-mode units is in the range of 23-25° API, and Conradson carbon residue (CCR) for the Coastal and Mapco units are 0.19 wt % and 2.7 wt %, respectively. Therefore, the feed quality of Indian Oil's FCC unit and Coastal's FCC unit are very similar, but feed quality of Mapco's FCC unit is heavier in terms of feed density and feed conradson carbon.⁵

Table 5 shows that the BCA reduces bottom yield irrespective of FCC mode of operation. Due to operating severity, the bottoms are upgraded to gasoline for gasoline mode and upgraded to TCO for distillate mode. It is interesting to note that production of fuel gas/dry gas has been reduced for all cases.

As expected, the reduction of regenerator temperature is also observed for all three cases.

Economic benefit

The cost of BCA at the Indian plant site is $4,620/ton. For the plant trial and for 1 year of operation, 10 tons of BCA were required, or $46,200/year. Based on the yield improvement with BCA during the test run and the international pricing of different products, the revenue generated with BCA is $1.54 million/year. Therefore, the net economic benefits is $1.50 million/year.

Based on the successful plant trial and the economic benefit, the Indian Oil refinery has been continuously using this BCA since 1995. Moreover, it has been found that the BCA minimizes bottoms while processing heavier VGOs, particularly those containing higher basic nitrogen and aromatics. The dosing rate of the additive and its concentration in the catalyst inventory is varied, depending on the vacuum gas oil (VGO) quality. Typically, the BCA concentration is 0-5 wt % on the catalyst.

Acknowledgments

The authors thank the management of Indian Oil for giving permission to publish this work. Thanks are also due to our colleagues at the Mathura refinery, particularly Mr. V.B. Shende, who was associated with laboratory experimentation.

References

1. Mitchell, Jr., M.M. and Goolsby, T.L., "Improvement of FCC Catalyst Performance with Bottom-cracking Additives," Symposium on New FCC Technology: Additives, 1990 Spring National AIChE Meeting, Orlando, Mar. 18-22, 1990.
2. Krishna, A.S., Hsich, C.R., English, A.R., Pecoraro, T.A., and Kuehler, C.W., "Additives Improve FCC Process," Hydrocarbon Processing, Vol. 70, No. 11, 1991, p. 59.
3. Mandal, S., Bhattacharyya, D., Shende, V.B., Das, A.K., and Ghosh, S., "Impact of Additives Usage in distillate Fluid Catalytic Cracking Operation," ACS Symposium Series 571, ACS, Washington, Chap. 24, 1994, pp. 335-48.
4. Ellison, T.W., Demmel, T.W., Steves, C.A., and Johnson, C.R., "Enhanced Bottoms Cracking Using A Solid Particle Additive in a Commercial FCC Unit," Fuel Reformulation, 1993(3), p. 18.
5. Evans, M., and Brown, M., "Commercial Experience with BCA-105 Bottoms Cracking Additive," Resid Upgrading Alternatives section of the 1996 Spring National AIChE Meeting, New Orleans, Feb. 26-29, 1996.
6. Smith, G.A., Santos, G., Hunkus, S., and Tucker, T., "Optimizing FCC Operations and Economics by Improving.Feed Conversion and Total Liquid Yield with BCA-105," 1994 NPRA Annual Meeting, Mar. 20-22, 1994, San Antonio.