You are currently viewing: Articles



Mar-2013

Predicting FCC unit performance with laboratory testing

Proper simulation of e-cat and reaction conditions is critical for modelling future FCC unit operations

KENNETH BRYDEN
Grace Catalysts Technologies
Viewed : 4060
Article Summary
Why do we do testing? We do testing because we want to predict in the lab what is going to happen in the field. Leo Baekeland, an entrepreneur and pioneer in the plastics industry, famously spoke of the importance of lab and pilot plant testing when he stated in his 1916 Perkin Medal acceptance speech, “The principle: ‘Commit your blunders on a small scale and make your profits on a large scale,’ should guide everybody who enters into a new chemical enterprise.” Conducting testing before commercial implementation reduces risk for a refiner. Examples of questions that pilot testing can answer include:
• What will be the effect of a potential feedstock change on yields?
• How will a new catalyst technology perform?
• Which catalyst technology is best for my operating goals?
• What effect will an additive have on my yield structure?

Figure 1 outlines what pilot testing intends to accomplish. On a lab scale, the goal is to match the complex processes occurring in a commercial FCC unit. In the unit, catalyst deactivates over a period of many weeks due to temperature, steam and contaminant metals. Commercial deactivation conditions are too slow to be practically copied in the lab, so an accelerated lab deactivation is done to generate a simulated e-cat to match the chemical and physical properties of the commercial e-cat. Bench-scale (ACE  or MAT) or pilot-scale (DCR) test equipment is then used to simulate the reaction conditions in the FCC unit and react catalyst and feed to produce products.

Laboratory deactivation approaches
When studies are being done for feedstock selection or process development, commercial e-cat is usually used and no lab deactivation is needed. However, catalyst selection studies and catalyst R&D start with fresh catalyst. Fresh FCC catalysts need to be deactivated before testing because fresh catalysts are too active and the selectivities seen in fresh catalysts do not represent e-cats in the FCC unit. Temperature, metals and steam are therefore used to turn fresh catalyst into simulated e-cat. Commercial e-cat properties that we want to match with simulated e-cat include: surface area, unit cell size (UCS), metals concentration, metals oxidation state and metals distribution.

Accelerated conditions to simulate hydrothermal deactivation of zeolite typically involve times of two to 50 hours, temperatures between 1400°F and 1525°F (760°C and 830°C), and steam concentrations between 50% and 100%.1 At temperatures below 1400°F, it may be impossible to match the equilibrium UCS, and at temperatures above 1525°F unrealistic zeolite sintering can be encountered.

Contaminant metals such as nickel and vanadium accelerate catalyst deactivation and have dehydrogenation activity that increases coke and hydrogen. It is important to test catalysts with contaminant metals in order to realistically assess the performance of catalysts with metals trapping and passivation technologies. The best way to simulate the contaminant metals is to apply the same metals level to the fresh catalyst that is present on the e-cat, and then deactivate all the fresh catalyst samples in a study under the same conditions.

Deactivation methods that simulate poisoning by contaminant metals include the Mitchell method (MM), cyclic metals impregnation (CMI) and cyclic propylene steaming (CPS).2 The Mitchell method involves impregnation of fresh catalyst with organic nickel and vanadium naphthenates followed by steam deactivation for four to 20 hours. The CMI method involves multiple cycles of cracking with metals spiked feedstock and regeneration, resulting in a deactivation time of more than 50 hours. The CPS method involves impregnation of fresh catalyst with organic nickel and vanadium compounds, followed by ageing in a cyclic redox environment for ~20 hours. The reducing atmosphere (which simulates the riser) is a blend of steam, nitrogen and propylene, and the oxidising atmosphere (which simulates the regenerator) is a blend of steam, air and SO2. Grace developed the CPS deactivation method, and additional details can be found in reference 3. The CPS method provides a good match to e-cat properties and yields, as seen in Tables 1 and 2, where the properties and pilot plant yields of fresh catalyst deactivated via CPS are compared to commercial e-cat. The CPS method has been adopted by many labs around the world and can be easily fine-tuned to match the severity and specific deactivation conditions of different commercial units by adjusting the temperature, the number of redox cycles and the amount of oxygen in the regeneration atmosphere.

Table 3 is a comparison of the different lab deactivation methods to the conditions in a commercial unit. In deactivating catalyst with contaminant metals, it is important to include the effect of sulphur competition by using SO2 as part of the simulated regenerator conditions.

Under commercial regenerator conditions, calcium oxide and barium based metals traps are rapidly poisoned by sulphur and lose their vanadium trapping ability. This sulphur poisoning does not happen with rare earth based vanadium traps. Testing of vanadium traps in the laboratory without simulating the SO2 present in a commercial regenerator can give a false prediction of catalyst metals trapping ability.

Commercial FCC units differ in their catalyst turnover rates. When it is desired to very closely match lab-simulated e-cat to e-cat from a specific refinery, age distribution deactivation can be used. Commercial e-cat consists of catalyst particles with varying age, surface area, UCS, metals level, activity and selectivity. Sink/float experiments that separated refinery e-cat into age fractions have determined that activity and selectivity are dominated by the youngest fraction of the catalyst. Typically, the youngest 20% of the inventory contains less than 10% of the contaminant metals and contributes about 50% of the overall activity. For a specific unit, the metals and activity distribution will depend on catalyst addition rate, deactivation rate and catalyst activity. Figure 2 presents the contaminant nickel and activity distribution for e-cat from a refiner on the US east coast and how this age distribution can be simulated by deactivating three separate catalyst fractions: one representing the youngest 20%, one representing the middle 20% and one representing the oldest 60%. A greater number of fractions could be used in the simulation, but Grace has found that three age fractions results in the least complexity while still giving a good match to commercial yields. Table 4 shows the chemical and physical properties of the three fractions that were deactivated, Table 5 shows how the properties of a blend of the three fractions match those of commercial e-cat, and Table 6 provides yields from the DCR evaluation of commercial e-cat and the simulated e-cat produced by the AD-1 protocol. As the DCR results show, there is excellent agreement between the yields of the e-cat and the AD-1 deactivated catalyst.
Current Rating :  3

Add your rating:



Your rate: 1 2 3 4 5