Worldwide Catalyst Report Experts reveal catalyst-selection methodologies

Fig 1. - FCC Catalyst quality control tests [38614 bytes]

Refining catalyst selection procedures were discussed in detail at Oil & Gas Journal's International Catalyst Conference, Feb. 1-2, in Houston.

Marathon Oil Co.'s James P. Wick said, "With increasing cost pressures and tight margins, it is more important than ever to maintain a high level of technical satisfaction with the catalysts utilized in a refining system, and to tightly manage such a high-impact and significant cost area.

"Failure to assure the use of optimum catalysts may not result in any overt cost or operating problem," said Wick, "but will be manifested in subtle or obvious yield value changes, constraint violations, and run length definitions."

Wick revealed details of Marathon's program for review and optimization of fluid catalytic cracking (FCC) and hydrotreating catalysts. And renowned FCC expert Del Tolen outlined a step-by-step procedure for choosing an FCC catalyst.

Marathon's program

Marathon's program balances opportunities and risks against cost issues, says Wick. The FCC unit (FCCU) provides one of a refinery's most prominent opportunities for economic gain.

For Marathon, economic drivers for FCC catalyst selection are particularly site-specific because the company's four FCCUs range in size from 28,000 b/d to 95,000 b/d. In addition, feed quality and unit operating objectives vary considerably.

Marathon refineries review FCC catalysts at least every 3 years, and sometimes more often if there have been:

Unit changes, including catalyst modifications, feedstock shifts, and model-based economic alterations
Catalyst developments
Vendor problems.

For an FCC catalyst study, Marathon supplies unit constraints and objectives to catalyst vendors, who then provide catalyst samples to Marathon's Petroleum Technology Center in Littleton, Colo. When a catalyst is chosen, the properties of the sample form the defining specification in Marathon's purchase agreement with the vendor.

The catalyst samples are steamed so as to mimic the current equilibrium conversion activity of the incumbent catalyst. Technicians then use refinery feedstock to perform microactivity tests (MATs) on the catalyst samples.

Each catalyst is run at three catalyst-to-oil ratios. At least three runs are performed at each ratio, and a 2.5% mass-balance closure limit must be met. In this manner, detailed data for yields, conversion, specific activity, and coke selectivity are collected.

"Interpolated yields at constant coke and conversion are compared as a base to the model work," said Wick. "The regressed MAT data are input into a commercial FCC model and the predicted refinery FCC yields and process conditions are defined. The predicted yields are then entered into our LP and a net value differential is determined for each catalyst."

Typically, several candidate catalysts will produce similar value-added estimates.

New catalyst performance is cross-checked with vendor data. If any of the physical or chemical properties do not support Marathon's findings, the discrepancies are investigated.

This process narrows the field of candidates, and catalyst cost is introduced as a key factor.

"An important component of cost control is development of supply contracts," said Wick. "These contracts contain specific, negotiated price-reduction components.

"Another beneficial feature of supplier contracts is specification definitions, with remedy clauses for off-spec material," he added.

Marathon ensures catalyst quality by analyzing composite samples monthly. Any deviations found are negotiated with the vendor.

Fig. 1 shows an example of Marathon's test results for two FCC catalysts.

Marathon also performs post-audits on any significant changes in FCC operations, including catalyst changes. These audits usually are performed 3-6 months after a catalyst change is initiated.

Actual yields are compared to predictions from: Marathon's MAT tests, model runs, and vendor data.

"In a typical Marathon study," said Wick, "the actual yield values were shown to be advantageous over the previous catalyst by $0.23/bbl, whereas the predicted advantage was $0.15/bbl."

To complete the loop, Marathon may alter the catalyst formulation or the model data.

Hydrotreating

Marathon's 12 hydrotreaters process feeds ranging from naphtha to heavy gas oil. Unlike FCCUs, similar hydrotreaters can share catalyst evaluation data.

Marathon categorizes its units as either high or medium impact. High-impact units include FCC pretreaters and some distillate hydrotreaters.

Marathon always uses fresh catalyst in its high-impact units, so the bulk of its hydrotreating catalyst evaluations are done for these units.

Catalyst recommendations are solicited from vendors 6-12 months before a scheduled catalyst selection date. The field is narrowed to three to seven choices, and the catalysts are tested in Xytel pilot plants at the Littleton laboratories.

Typically, technicians run identical feedstocks in four separately controlled pilot units. Kinetic activity is defined at a variety of hydrotreating severities.

The tests are perturbed with extreme feed types, especially at the heavy end, says Wick. The purposes of introducing this upset are to increase the performance differential between catalysts, and to determine a catalyst's ability to recover after a feed upset.

Aging tests are run on some catalysts to determine deactivation rates, but these tests usually are run for feed studies rather than catalyst tests.

As with its FCC catalysts, Marathon also analyzes the physical and chemical properties of candidate hydrotreating catalysts.

To reduce hydrotreating catalyst costs, Marathon purchases catalyst for several similar units from a single vendor at one time.

"This allows the vendor to better plan his production runs at lower cost, which we share," said Wick.

Marathon also keys these large purchases to metals prices. The company has even purchased low-price metals and supplied them to its catalyst vendor.

For quality control purposes, Marathon samples the "as-charged" load of hydrotreating catalyst. The samples are tested only if unit performance is problematic.

"Very infrequently have we seen quality assurance problems," said Wick. "The most predominant problem appears to be control of physical parameters such as short extrudates, not with the activity of the purchased load."

Marathon's fixed-bed catalyst post-audits center on feed and severity optimization. Occasionally, activity is suspect and retained samples are evaluated.

Key to Marathon's hydrotreating catalyst management system is regenerating and reusing its catalysts, says Wick.

"Specific high-impact units within our system are the low-metals 'suppliers' of active, top-of-the-line regenerated catalysts."

The used catalysts are regenerated through an exclusive contract with a commercial regenerator, who stores the catalysts until Marathon needs them. The regenerated catalysts are then used for lower-severity service.

"Our cascading system is nearly balanced," says Wick, "so we do not need to purchase regenerated non-Marathon catalysts."

Marathon's system of catalyst selection and optimization has enabled it to extend run lengths in many of its fixed-bed units to 5 years from 4 years.

Tolen's selection process

According to Del Tolen, switching to a more favorable catalyst can improve profits in a refinery with a 10,000 b/d FCCU by $3-5 million/year. Tolen, who heads Rocky Mountain Salvage & Equipment Inc., Grand Junction, Colo., presented a detailed method of choosing an FCC catalyst.

Tolen's method is based on how the catalysts have performed in operating units. It takes into account variations in unit configuration, operations, and product requirements.

"Similar units with similar feeds give the best comparisons," said Tolen, "but lacking those, similar feeds are most important."

Catalyst vendors are an important source of information, says Tolen, but other refiners, unit operators, and engineers are a more important data source. For this purpose, a number of good contacts can be made at industry meetings.

If a refinery is one of many within a company, engineers in affiliate refineries can be consulted. Process licensers, consultants, and supplier seminars also are potential sources of contacts and operating data.

In addition, catalyst suppliers should be able to provide a list of 10-12 other refineries with similar feeds and configurations.

Data gathering

In order to standardize the data collected, Tolen suggests using forms like the ones shown in Tables 1 [.pdf file] and 2 [42785 bytes]. These forms will help refiners record and sort considerable data concerning operating conditions, yields, and unit configuration.

"The operating conditions and yield sheet (Table 1) will allow the engineer to discern how another unit is run, the feed and product qualities, and the volumetric and weight yields," said Tolen. The unit configuration and limits form (Table 2) provides details about mechanical configuration and unit style.

"For instance," said Tolen, "if there is a limit on stack opacity or total solids emissions, these should be listed on the limits sheet."

Tolen says the most difficult comparisons to be made are those concerning feed quality. While many evaluation methods have been proposed over the years, none are as accurate as the results from an operating FCCU.

Feed quality

"The feed charged to the unit is the most important variable in determining yields, and thus in choosing a catalyst," says Tolen. "In my experience, the best tests for determining feed quality are refractive index, aniline number, and basic nitrogen."

For resids, the first two qualities are difficult to determine by traditional methods. A thin-film refractive index and dilute-method aniline number test are available for this purpose.

When choosing a catalyst, the goal is to get the greatest quantity of high-value products from the lowest-cost feed, thus achieving the highest possible margin. One important feed parameter is gasoline efficiency, defined as the volume of gasoline yield divided by the conversion (in other words, the gasoline yield per unit of conversion).

Tolen's recommendations for matching catalyst with feed are:

For light, paraffinic feeds, a high-zeolite catalyst with good strippability, adequate hardness, and low gas make.
For mixed gas oils in the medium boiling range, a more open-pore structure with a high zeolite content and some matrix activity.
For heavy gas oils and resid feeds, a catalyst with lower zeolite content, considerable matrix activity, and a very open pore structure.

A high matrix activity cracks heavy components, binds nickel, and reduces gas yields. Because some catalysts purported to have high matrix activity do not live up to those claims, Tolen suggests a test method for determining "real" matrix activity: Steam a sample of fresh catalyst at 1,500° F., test it for remaining activity, and compare the results to a catalyst with mostly zeolite surface area.

A high pore volume also increases cracking of heavy components. In addition, it increases penetration, absorption, and zeolite access for small molecules. Increased pore volume will boost strippability and reduce coke make, catalyst deactivation, and regenerator NOx formation.

Tolen warns, however, that high-pore-volume catalyst has a higher attrition index.

Selection method

Once the data from other refineries have been gathered, Tolen recommends the following method for narrowing down the choices:

1. Sort the feed analyses by refractive index and aniline number. Include at least four to six feeds that are within 0.05 refractive index numbers from your feed.

2. Sort again, taking those that are within 15° F. in aniline number. Three to six feeds should remain from the starting group.

3. If desired, use UOP K factor to further eliminate discrepancies.

4. Sort again, choosing data from units with speed and degree of mixing similar to your own.

5. Next, consider riser injection technology and, particularly, feed nozzle design. These factors affect mixing, which, in turn, influences dry gas make, C3/C4 make, C3/C4 olefinicity, and total coke make.

6. Sort by feed nozzle velocity.

7. Continue sorting by the limiting FCCU parameters (catalyst loss, C3/C4 yield, coke make, regenerator temperature). In some cases, stack opacity limitations may make catalyst hardness or density a primary concern, and this should be considered.

8. Compare other refiners' results when operating on the sort of heavy, dirty feeds you would like to run.

9. Send feed samples to three or four catalyst suppliers. Include a feed "pedigree" (this should name the crude or crudes from which the feed is derived, list other streams included, and state whether the feed is hydrotreated and at what pressure, etc.).

10. Ask the supplier to recommend two catalysts and provide yield estimates for each.

11. Obtain from the supplier a list of refiners that are using the recommended catalysts.

12. Check all yield estimates for weight and hydrogen balance; discard any that are in error by more than 2.0%.

13. Obtain operating results from other refiners using these catalysts and request their comments about the supplier's customer service.

14. If, for the remaining catalysts, yield estimates are reasonable, service is comparable, and your intended use fits the supplier's recommendations, use the supplier's yield estimates in your economic calculations and choose the optimum catalyst.

When performing this procedure, take into account whether a certain catalyst is used often for a particular service.

"For instance," said Tolen, "does one particular catalyst brand predominate where octane is required, where bottoms cracking is maximized, or where high isobutylene yields are needed?"

If so, the catalyst choice should be skewed toward that catalyst if the economics are close.

"Catalyst decisions should be based on maximum input from other refiners, and on all the hard data that can be garnered from them," said Tolen. "Particular attention should be paid to getting the catalyst suppliers to define the capabilities of their various catalysts."

And, finally, if you discover that you've made an error, do not hesitate to switch to the second catalyst on your list.