SO x control additive reduces emissions, saving capital investments

Dec. 29, 1997
SO x control additives can reduce fluid catalytic cracking (FCC) SO x emissions to comply with environmental regulations without requiring the capital investment and maintenance cost of feed hydrodesulfurization or flue-gas scrubbing. Two cases described in this article reveal the additives' capabilities and cost. Citgo Petroleum Corp., in a full-burn test in its Lake Charles, La., refinery, achieved an SO x reduction of 44% when testing Grace Davison's Desox SO x
John S. Becker
Ultramar Diamond Shamrock Corp.
Alma, Mich.

Mark J. LaCour
Citgo Petroleum Corp.
Lake Charles, La.

Steven W. Davey, John T. Haley
Grace Davison
Baltimore

SO x control additives can reduce fluid catalytic cracking (FCC) SO xemissions to comply with environmental regulations without requiring the capital investment and maintenance cost of feed hydrodesulfurization or flue-gas scrubbing.

Two cases described in this article reveal the additives' capabilities and cost.

Citgo Petroleum Corp., in a full-burn test in its Lake Charles, La., refinery, achieved an SOxreduction of 44% when testing Grace Davison's Desox SOxcontrol additive. Total Petroleum (North America) Ltd. (now Ultramar Diamond Shamrock Corp.), in a partial-burn test in its Alma, Mich., refinery, achieved an SOxreduction of 38%.

FCC SOxemission

Sulfur present in the FCC feed is distributed into the products in various quantities. If available, downstream hydrotreating can reduce the sulfur in the liquid products while H 2S in the off gas is removed in the sulfur-recovery unit. The sulfur in the coke deposited on the catalyst leaves the FCC process as flue gas from the regenerator in the form of SO 2 and SO 3.

Since Desox's commercialization in 1986, over 70 refineries have used it to control flue gas SOxemissions.

FCC process and sulfur

In the FCC process, feed is injected into the riser section of the reactor where it mixes with catalyst and is cracked as it passes up the riser. The products are then separated from the catalyst by reactor cyclones.

The catalyst, now covered with coke, is stripped of hydrocarbons by steam and then directed to the regenerator. There, the catalyst is regenerated by contact with air, which burns off the coke.

The regenerated catalyst repeats the cycle. The combustion products (CO2, CO, SOx, and NOx) produced from the coke burning leave the regenerator as flue gas.

Table 1 [59,248 bytes] summarizes typical sulfur distributions in FCC products. Sulfur in coke contains 2-30 wt % of FCC feed sulfur.

The sulfur distribution is dependent on the species of the sulfur contained in the feed and the extent of feed pretreatment or hydrotreating used by the refiner (OGJ, Aug. 2, 1971, p. 64 and OGJ, May 19, 1975, p. 73).

SOxtransfer mechanism

In the FCC process, catalyst circulating between the oxidizing atmosphere of the regenerator and the reducing atmosphere of the reactor makes the transfer of sulfur via an SO xcontrol catalyst possible. The additive traps SO xin the regenerator and releases it as H 2S in the reactor. The H 2S released in the reactor effluent results in a 5-20% increase in total H 2S, which can be handled in an existing gas concentration plant.

Desox catalytic SOxcontrol is described by the three-step process illustrated in Fig. 1 [58,664 bytes]. In Step 1, the sulfur in coke is oxidized to SO2, which further reacts with oxygen in the regenerator bed to form SO3. In Step 2, the sulfur is absorbed in the form of SO3 by reacting with the additive to form a stable metal sulfate. In Step 3, the sulfate is reduced in the reactor either directly to H2S or to a metal sulfide which is subsequently converted into H2S by steam in the reactor or stripper.

FCC operating conditions for optimal Desox performance include good regenerator air distribution, high catalyst-to-oil ratio, and high reactor stripping efficiency.

Efficient air and catalyst mixing impacts Desox usage rate. Other aspects of FCC design that influence Desox performance include extent of CO combustion, regenerator oxygen concentration, and stripping steam rate.

In commercial experience, units with a low catalyst inventory and a high catalyst circulation rate generally have high Desox pickup factors. This experience implies that Desox performance is enhanced by increasing the rate of contact with available sulfur, which correlates with a high rate of sulfate absorption in the regenerator.

Full-burn test

Citgo's Lake Charles refinery "A" FCC unit is an M.W. Kellogg Technology Co. design with a 500-ton catalyst inventory. It operates at a feed rate of about 37,000 b/d and a catalyst-to- oil ratio of about 6.0. The typical feed contains vacuum and coker gas oils with a gravity of 21.2° API and 2.3 wt % sulfur. The reactor outlet temperature is 985° F. The regenerator operates in full burn with a dense bed temperature of 1,330° F. Combustion promoter use is sporadic.

A baseline for FCC unit SOxemissions was established prior to Desox testing. The baseline allows the refiner to calculate the SOxemissions when Desox is in the circulating inventory. This calculated number is used to estimate the effectiveness in the unit.

Before the introduction of Desox into the Lake Charles FCC unit, a regression of operating data produced the following base emission correlation:

Flue gas sulfur (%ffs) = Slurry sulfur (%ffs) x 0.1862 + Slurry gravity (°API) x 0.0034 - 0.0162, where ffs is fresh feed sulfur.

The form of this equation is typical for most FCC units.1 2 Slurry sulfur is a good predictor of thiophenic feed sulfur which is commonly considered an excellent predictor of coke sulfur or flue gas sulfur.

After the base SO2 emissions have been estimated, the amount of SO2 removed during Desox use is calculated by subtracting the actual SO2 emissions at steady state from the calculated base SO2 emissions. Desox performance is then determined based on the amount of SO2 removed relative to the additive usage rate.

Since this evaluation was not run at steady state, the data must be modeled to determine the actual effectiveness of Desox. Performance modeling takes into account changes in operating conditions and variations in additive usage rates. It allows calculation of initial activity and decay characteristics vs. the addition rate. The general form for Desox performance modeling is:

Activity = Initial activity x e -(decay rate x time)
Results from Citgo's modeling are summarized in Figs. 2 and 3. Fig. 2 [82,452 bytes] shows the calculated base flue gas SO 2 emissions from the base emissions equation. It also shows the measured SO 2 emissions, model predicted SO 2 emissions, and Desox usage rate.

Fig. 3 [90,576 bytes] shows the flue gas SO2 removed and the Desox usage rate.

Once the initial activity and decay rate coefficients for Desox have been determined by modeling, the steady state performance of Desox and the cost to control SOxemissions can be calculated. The economic evaluation is shown in Table 2 [5,585 bytes].

To achieve environmental compliance, the Lake Charles refinery was required to remove 8,100 lb/day of sulfur from the FCC stack. During the evaluation, an average 410 lb/day of Desox was used to reduce flue gas SO2 emissions by 13,120 lb/day, yielding a pickup factor of 32 lb of SOxremoved per pound of Desox used per day. The average cost per pound of sulfur captured is $0.24.

Partial-burn test

Ultramar Diamond Shamrock's Alma FCC unit is a UOP stacked design with a 60-ton catalyst inventory. It operates at a feed rate of about 20,000 b/d and a catalyst-to-oil ratio of about 4.9. The typical feed is a mixture of reduced crude and hydrotreated feed with a gravity of 28.6° API and 0.8 wt % sulfur. The reactor outlet temperature is 995° F. The regenerator operates in partial burn with a dense bed temperature of 1,410° F. The refinery uses CP-5 CO combustion promoter to periodically control afterburn.

Each FCC unit that tests Desox requires a base emissions correlation to accurately determine FCC unit SO2 emissions. Since the Alma FCC unit began Desox use coincident with a process unit startup, no base emissions correlation is available. Through modeling Alma emissions data, a base emission correlation was developed that assumes a linear relationship between flue gas sulfur and slurry sulfur, both as a percentage of fresh feed sulfur:

Flue gas sulfur (%ffs) = 0.50 x Slurry sulfur (%ffs)
The Alma FCC unit has used catalytic SO xcontrol since April 1994. Through operating improvements and catalyst selection, it has significantly reduced delta coke. The lower delta coke operation resulted in a cooler regenerator. This freed up air blower capacity, which moved the FCC unit from a deep partial-burn (CO 2:CO ratio of 1.0) to a more typical partial-burn operation (CO 2:CO ratio of 4.0). Even a slight increase in the availability of oxygen in the regenerator bed had a significant impact on Desox performance.

Fig. 4 [87,860 bytes] shows the performance of Desox in the Alma FCC unit from a recent test run. It shows SO2 removed and the corresponding Desox addition rate.

As with the full-burn example, we used a model to predict Desox performance at steady state. The results of the partial-burn modeling are in Table 3.

The Alma refinery removes 1,897 lb/day of sulfur from the FCC stack. Based on these test results, 179 lb/day of Desox was required, which yields a pickup factor of 12 lb of SOxper pound of Desox used. The average cost per pound of sulfur capture is $0.74.

Performance comparison

The SO xcontrol catalyst in the full-burn test was 2.7 times more effective than in the partial-burn test.

Generally, in a given unit, Desox will perform better in full burn than in partial burn. The critical step in capturing flue gas sulfur is the oxidation of SO2 to SO3. Although Desox technology incorporates a catalyst to drive the oxidation reaction, operating with excess oxygen in the regenerator will favor SO3 formation. Actually, any operating changes that increase the oxidizing environment of the regenerator, including the use of CO oxidation promoter, will improve the performance of an SOxcontrol catalyst.

Several refineries, however, have reduced emissions by 50% or more while operating in partial burn.2

Controlling SOx

The application of catalytic SO xcontrol with Desox is not limited to control levels of 40-50%. Testing at a West Coast refinery confirmed that Desox can be used to reduce SO xemissions by 95%.3 In this test, Desox was added to achieve a proposed South Coast Air Quality Management District target of 6 kg of SO xper thousand barrels (6 kg/1,000 bbl) of FCC feed. The refiner was able to reduce SO 2 emissions to 10 ppm with the addition of 325 lb/day of Desox. The pickup factor during the test was 7, and the Desox concentration in the circulating inventory during the test was 8%. This test and others confirm that Desox can be used to cost-effectively control SO xto very low levels.

Flue gas scrubbing comparison

The New Source Performance Standards (NSPS) of the Environmental Protection Agency (EPA) went into effect in 1989. They cover new FCC units or modified units since January 1984. Refiners impacted by NSPS have four options to reduce FCC flue gas SOx: flue gas scrubbing, low sulfur feedstock processing, hydrotreating, and catalytic SO xcontrol.

If a refiner selects flue gas scrubbing, SOxemissions must be reduced by 90% or emissions must be kept below 50 ppm. A flue gas scrubber requires a significant capital investment ($16 million to $20 million including a purge treatment unit), has ongoing operating costs (caustic and maintenance), and presents waste disposal problems.

The economics of flue gas scrubbing vs. catalytic SOxcontrol with Desox can be compared as follows.

The basis for comparison is a typical 50,000 b/d FCC unit processing 25° API feed with 19.6 kg flue gas SO2 emissions per 1,000 kg coke (about 600 ppm SO2). Under NSPS, using an SOxcontrol additive will require this refiner to reduce emissions by 50% to 9.8 kg SO2/1,000 kg of coke burn. The operating cost of this option, based on standard Desox performance, is $3,500/day.

If a refiner installs a flue gas scrubber, the SO2 emissions must be reduced by 90%. The operating cost for the flue gas scrubbing option is estimated to be $8,000 per day, excluding amortization of the capital cost of the scrubber.

The break-even point for flue gas scrubbing vs. catalytic control is estimated to be over 2,500 ppm SO2 in the FCC flue gas. Because most refiners have uncontrolled SO2 emissions well below this level, the number of refiners operating flue gas scrubbers worldwide is small.

References

  1. Powell, J.W., et al., "Reducing FCC Emissions with No Capital Cost," 1990 AIChE Spring National Meeting, Orlando, March 1990.
  2. Latimer, J.A., et al., "Controlling FCC SOxEmissions with Desox, Effective Performance in Partial Combustion," 1991 AIChE Annual Meeting, Los Angeles, November 1991.
  3. "Reducing FCC SOxEmissions to Very Low Levels Using Desox," UOP, 1992 NPRA Annual Meeting, New Orleans, March 1992.

The Authors

John S. Becker is the FCC superintendent at the Ultramar Diamond Shamrock refinery in Alma, Mich. He has 7 years of operations experience in catalytic cracking and an additional 5 years of experience in refinery design, project maintenance, and operations engineering. He holds a BS in civil engineering from Michigan Technological University
Mark J. LaCour is the unit supervisor for the "A" FCC unit and the propylene fractionation unit at the Citgo Petroleum Corp., Lake Charles, La., facility. Previously, he worked as the FCCU engineering supervisor for Citgo. LaCour also has been a Grace Davison chemical technical service representative and a process engineer on the FCC unit at Murphy Oil Corp., Meraux, La. LaCour holds a BS in chemical engineering from Louisiana State University and an MBA from the University of New Orleans.
Steven W. Davey is the director for the FCC additives business at Grace Davison, where he has worked for the past 16 years in various positions, including technical service, marketing, finance, and process engineering. Davey holds a BS in chemical engineering from the University of Delaware and an MBA from Loyola College, Baltimore.
John T. Haley is the technical supervisor for Grace Davison's FCC additives group. He has over 12 years of experience in fluid catalytic cracking. Previously, he worked in various positions at Katalistiks, a former FCC catalyst company. Haley holds a BS in chemistry from Towson State University, Maryland.

Copyright 1997 Oil & Gas Journal. All Rights Reserved.