Oct-2015

Performance testing of naphtha 
reforming catalysts

High throughput screening of naphtha reforming catalysts under commercially relevant conditions

MARIUS KIRCHMANN, ALFRED HAAS, CHRISTOPH HAUBER and SASCHA VUKOJEVIC, hte GmbH

Article Summary

Catalytic naphtha reforming, in which low octane naphtha feedstock is converted into high octane reformate, is one of the core processes in modern refineries.^1,2 The reformate produced includes high-value aromatics for the petrochemical industry such as benzene, toluene, and xylenes (BTX). Hydrogen as a main byproduct is highly valued for its use in hydrotreating, hydrocracking, and other hydrogen consuming processes in the refinery.

There has been ongoing research in the last decades with the aim of optimising activity, selectivity and stability in order to increase high octane C5+, aromatics and hydrogen yield. More recently, additional challenges have emerged due to environmental regulations, requiring a reduction of aromatics content (especially benzene) in gasoline and an increase in hydrogen for the production of clean fuels.

On the other hand, market trends show increasing global demand for aromatics in the petrochemical industry. In regions of developing markets such as Asia, strong growth in gasoline demand is expected, while demand is declining in developed regions such as North America, Europe and Japan.³

Refiners need to decide whether to meet gasoline or aromatics demand by changing the catalyst, changing the mode of operation, revamping existing reformers from fixed bed to continuous catalyst regeneration (CCR) reformers, or installing new reforming capacity in regions close to developing and growing economies. At the latest when a catalyst reaches the end of its service life, the difficult question comes up of whether to reduce risk and stick to the same catalyst, or to select a new, potentially better performing catalyst for the change-out. Besides risk minimisation, catalyst costs have to be taken into account versus increase in profitability, operability, stability and the capability to deal with different feed compositions. This is made more difficult by the fact that commercial naphtha reforming catalysts on the market have been optimised for decades and differences in performance can be very small. Notwithstanding this, even small differences in performance have a great impact on process profitability due to the large capacity of reformers. Therefore, an independent catalyst test to benchmark catalysts on the market and compare their performance to the catalyst currently in service should be considered to minimise risks, increase economic return and help to make the right decision.

Detecting these small differences in performance, however, takes conventional sequential catalyst testing in single-fold pilot plants to the limit and even minor deviations in process variables and the calibration of technical equipment (temperatures, pressures, flows, and analytics), feed composition or catalyst ageing between runs can easily compromise the results. Multiple runs have to be performed in order to obtain adequate statistical significance for reliable results, in the worst case not yielding sufficiently small error limits or reproducibility to differentiate between catalysts. In addition, running these tests for multiple catalysts or process conditions consumes time that is often not available and is expensive.

High throughput experimentation (HTE) technology can increase efficiency by testing many reactors in parallel, which allows several catalysts and process conditions to be tested simultaneously and saves time and costs. The option of installing the same catalyst in multiple positions immediately generates results with meaningful statistical significance. Nevertheless, differentiation of catalysts is only achieved if performance differences are higher than the statistical error of measurements, thus rendering catalytic naphtha reforming a challenging application for parallel testing of catalysts. Reproducible reactor loading, constant process conditions in each reactor (temperatures, pressures, flow distribution) and high analytical precision are of great importance.

Catalytic naphtha reforming is predominantly carried out in fixed bed reformers comprising both cyclic and semi-regenerative operations, though newer and more effective CCR reformers have high penetration nowadays and are continuing to gain ground. Process conditions and catalysts for these processes differ significantly from each other and require a highly flexible parallel test unit to realise both protocols. The long cycle length and slow deactivation of fixed bed reforming catalysts (up to a year or more) requires a fast testing approach employing accelerated decay conditions in order to gather results in a reasonable time frame. On the other hand, short deactivation of CCR reforming catalysts demands a fast approach in order to collect enough data before the catalyst deactivates, by reducing analysis time to increase sampling frequency, increasing the parallelisation degree of the analytics, reducing the number of reactors that are on stream in parallel, or sequentially starting up single reactors.

In earlier stages of catalyst development with only small amounts of catalyst available as powder, testing under isothermal conditions can deliver initial qualitative results, identify leads, and provide a tool for fast performance checks after regenerations or to monitor production. Commercial naphtha reformers, however, operate adiabatically with multiple sequential reaction zones and inter-heaters in between (see Figure 1). Many types of reactions interact with each other, including desired reactions such as dehydrogenation, dehydrocyclisation or isomerisation and undesired reactions such as hydrocracking and hydrogenolysis. Especially in the first reactor of a commercial set-up, fast and endothermic reactions such as dehydrogenation and dehydrocyclisation prevail over exothermic reactions, leading to a significant temperature drop. If a catalyst is very active, it will cause a higher endothermic temperature decrease than less active catalysts. When benchmarking these catalysts under isothermal conditions, the relative rates and contributions of the reactions involved will be different, leading to different selectivities and in the worst case an incorrect ranking of catalysts. On the other hand, testing catalysts adiabatically in multiple sequential reactors requires an extensive set-up and high amounts of catalyst and feed even in a single-fold unit. In a 16-fold parallel test rig, a good compromise between complexity and closeness to industrial practice was found by multiple reaction zones that allow the development of semi-adiabatic temperature profiles with same weighted average inlet temperatures (WAITs). Finally, usage of full extrudates improves reproduction of transport phenomena such as heat and mass transfer and is closer to practical conditions.

In naphtha reforming, low octane paraffins and naphthenes in naphtha are converted to high octane iso-paraffins and aromatics in reformate. Along the cycle time, catalyst deactivation occurs predominantly through carbonaceous deposits and leads to a gradual decrease of the research octane number (RON). However, refineries rely on a constant reformer output for gasoline blending purposes, aromatics downstream processes or hydrogen supply. In order to compensate the RON decrease, the temperature can be increased or pressure and weight hourly space velocity (WHSV) decreased. While lowering pressure increases coking and shortens catalyst lifetime significantly (which makes it a good measure to accelerate decay in laboratory testing), decreasing WHSV limits capacity and increases hydrocracking to lights. Therefore, temperature remains as a sensible variable and operation at constant RON in fixed bed reformers is achieved by gradually increasing the temperature until heater capacity is reached or process economics become unfavourable due to increased hydrocracking and strongly reduced C5+ yields. In a parallel test rig, this constant or iso-RON operation requires a fast and reliable analytical method to determine RON in each reactor combined with individual temperature control for each reactor. In summary, the combination of multiple reaction zones with semi-adiabatic temperature profiles, usage of full extrudates, and operation under iso-RON conditions brings parallel testing very close to commercial practice.