Aug-2013

The importance of using pilot plant testing for FCC catalyst selection

In the EMEA region, over 65% of refineries perform catalyst testing prior to selecting FCC catalysts for their units.

Michel Melin, Stefan Brandt and Colin Baillie
Grace Catalysts Technologies

Article Summary

Many refining companies perform internal catalyst pilot plant testing, for example Total (an example of which will be duly discussed), Abu Dhabi Oil Refining Company (TAKREER) and Aramco. Many other refining companies perform external catalyst pilot plant testing at independent laboratories, including BP, Orpic and KNPC amongst them. In fact, BP concluded that post-audits of these commercial changes confirmed that they delivered significant value of several million dollars per year to the refineries.1 There are clearly large benefits for refining companies to perform catalyst testing, as discussed below, the most important being that pilot plant testing minimizes the financial risk associated with unsuccessful catalyst trials.

Why perform laboratory testing?
Performing a catalyst evaluation based only on paper studies is a risky strategy for refiners due to the possibility of inaccurate yield predictions, not subsequently realised in the FCC unit. Another alternative to laboratory testing is to conduct a direct plant trial, however there are also significant risks associated with this approach. Firstly, a fair comparison of two different catalysts is only possible when the feedstock quality and unit operating conditions are the same for both periods, something very rarely seen in practice. Even with the use of advanced process modelling software, it is difficult to compare catalyst performance during two different operating periods. This is particularly relevant when changes in hardware performance occur, such as stripper and feed nozzle deterioration. In addition, to be able to make conclusions from a plant trial requires a significant change out in catalyst inventory, typically over 60%. If there are undesired changes in the product yield pattern then this lengthy interval of sub-optimal catalyst performance will result in a large downturn in economic performance. During this change-out timeframe there is also the danger of unexpected impacts on FCC unit equipment, such as expanders. A catalyst selection study using pilot plant testing facilities is the only way to prevent these problems, and evaluate different catalyst technologies on a comparable basis. In addition to the importance of lab testing of catalyst yields, lab testing is also important in comparing the physical properties of different catalyst technologies. Different suppliers use different test procedures for measuring attrition that give different values. Testing the attrition performance of catalyst samples in the same facility gives a realistic performance comparison.

Catalyst deactivation procedures
The accurate evaluation of FCC catalyst performance in the pilot plant requires suitable procedures for the metallation and deactivation of fresh FCC catalyst to replicate the properties of FCC equilibrium catalyst (e-cat). There are two main deactivation mechanisms for FCC catalysts, metals deactivation and hydrothermal deactivation. For metals deactivation, Mitchell-type techniques2 are typically applied to impregnate FCC catalysts with vanadium and nickel. Recently, Grace has made an advancement in metals deactivation with the introduction of a new procedure that deposits metals on to the FCC catalyst particle using a spray drying technique.3 This gives a metal distribution that is more comparable to that of e-cats, where in particular nickel is enriched on the outer shell of the catalyst particle, as shown in Figure 1. For hydrothermal deactivation, Grace uses Cyclic Propylene Steaming (CPS) deactivation, which involves ageing the FCC catalyst in a redox environment for 20 hours. The advantages of using metals impregnation followed by CPS deactivation compared to alternative Cyclic Deactivation Unit (CDU) techniques is that it provides increased accuracy with respect to achieving the target metals levels, as shown in Figure 2. Additional advantages of this very robust method are that it matches e-cat activity and yields well, provides an accurate deactivation of metals traps, and is easier to tailor to match specific units. Consequently, the CPS methodology has found widespread acceptance in FCC catalyst evaluation, and is used in more than 60% of testing laboratories.

Reproducing diffusivity properties of e-cats
Diffusion limitations in FCC catalysts can be measured by pore volume. Figure 3 compares the pore volume of a Grace e-cat with the corresponding fresh catalyst deactivated by CPS, and it can be seen that the diffusivity properties of e-cats is very similar to the corresponding laboratory-deactivated catalysts.

Increased strippability of FCC catalysts can provide lower delta coke, which may not be reflected in laboratory testing. Here the concept of the SA/K number4 is important, as it indicates the amount of surface area (SA) required to give a unit of conversion. A lower SA/K number means that less surface area is required to achieve a certain activity, so for a given activity less surface area needs to be stripped. Figure 4 shows e-cat benchmarking of SA/K numbers for EMEA refineries, and shows that for a given unit cell size (UCS) Grace catalysts operate at lower SA/K numbers than those from other suppliers. These lower SA/K numbers (and more importantly the incorporation of industry-leading metals passivation technology) provides Grace customers with better coke and dry gas selectivities, as shown in Figure 5.

Commercial examples
The first example is based on a catalyst selection for an FCC unit in the EMEA region processing a desulphurised VGO feedstock, low in contaminant metals, and with the objective of maximizing propylene yield. The refinery tested two different catalyst technologies from Grace. Catalyst A has a low matrix surface area (MSA), and Catalyst B has a high MSA. Both catalysts were deactivated using the CPS procedure without metals impregnation, and evaluated using both the ACE^® evaluation unit and DCR^TM pilot plant testing with refinery feedstock. Figure 6 contains ACE^® evaluation unit, DCR^TM pilot plant and FCC unit data highlighting that the high MSA Catalyst B provides higher hydrogen yield and delta coke compared to the low MSA Catalyst A. Both sets of pilot plant data are clearly very much in line to the observed FCC unit data. Table 1 shows that the overall FCC unit yield structure could be very closely predicted using ACE^® evaluation unit and DCR^TM pilot plant testing, with the latter being particularly accurate.

Deactivation coefficients for zeolite and matrix are indeed different. However, it can clearly be seen that an accurate differentiation of performance between low and high zeolite-to-matrix (Z/M) catalysts can be made using pilot plant testing.

The second example of pilot plant testing being used to successfully predict FCC unit yields is taken from Total Antwerp, which is a resid unit with e-cat Ni+V levels of up to 10,000 ppm. Within the Total refineries, all new FCC catalysts must be qualified through pilot plant testing at the Total Research & Technology Gonfreville (TRTG) centre. Before trialling the Nektor catalyst from Grace at their Antwerp refinery the TRTG center performed an evaluation to compare the Nektor catalyst with the previous Kristal catalyst. Both fresh catalysts were deactivated in the laboratory, and Table 2 shows that very similar physical properties were obtained compared to the respective e-cats5.