December 2022

Special Focus: Catalysts

Solving maldistribution: Catalyst loading and why it matters

Are your reactor yields lower than forecast? Does your fixed-bed catalytic reactor suffer from hotspots?

Maas, E., Sigaud, J., Shell Catalysts & Technologies; Poussin, G., Caldas, T., CREALYST-Oil

Are your reactor yields lower than forecast? Does your fixed-bed catalytic reactor suffer from hotspots? Are your cycle lengths shorter than planned? If you answered “yes” to any of these, then your reactor could be suffering from liquid/gas maldistribution—one of the most common causes of poor reactor performance.

The impact of maldistribution can be severe and long-lasting, yet its root cause is often misunderstood. Although faulty or poorly installed reactor internals are often to blame, maldistribution can be frequently traced back to poor catalyst loading.

The authors’ companies collaborated to share insights into the impacts of maldistribution and the critical role that catalyst loading plays in reactor performance, as well as the important steps that plant personnel can take to ensure optimal performance of expensive catalyst and hardware.

Maldistribution: How is it affecting production?

Maldistribution is described as the preferential flow and the uneven distribution of liquids and/or gases within a catalyst bed, which can have the following consequences:

  • Off-spec products due to the suboptimal use of the loaded catalyst volume
  • Shorter-than-expected cycle length due to end-of-run (EOR) conditions reached sooner than expected
  • Hotspot formations and the resulting threat to the mechanical integrity of the reactor.

Performance impacts: Poor use of catalysts. Ideally, the weight hourly space velocity (WHSV) should be uniform within the entire catalyst bed volume; however, if a reactor is suffering from maldistribution, then portions of expensive catalyst are being underused (with a lower-than-average WHSV), resulting in lower-than-expected yields and off-spec products. Consequently, the reactor must run under more severe conditions to meet the targeted product properties or yields.

Time impacts: Shorter run/cycle times and increased downtime. In the most severe cases, maldistribution can lead to the early termination of the reactor cycle. This is caused by the development of hotspots—sections of the catalyst bed where the liquid/gas flow is impeded, causing high temperatures—that result in EOR temperatures being reached sooner than scheduled. Should a cycle be cut short, valuable run time is replaced with the costly and time-consuming process of replacing exhausted catalysts. In the most extreme cases, the hotspot temperature can be high enough to threaten the mechanical integrity of the reactor and cause severe safety incidents [e.g., reactor rupture, fire, explosion or hydrogen sulfide (H2S) release].

For example, imagine that plant personnel have targeted a 1,000-d unit cycle length, but the development of maldistribution forces the cycle to terminate after only 730 d, thereby reducing cycle run time by 27%. Stopping the unit earlier than scheduled and taking up to 3 wk to load a fresh batch of catalyst can impact the bottom line by as much as $18 MM (the cycle scenario assumes a planned yield quota of $500,000/d, a total planned cycle revenue of $500 MM, and replacement catalyst and downtime costs of $18 MM).

Cost impacts: Lost revenue and poor return on investment. Maldistribution results in lower yields, shorter cycles and more downtime, which, together, have a significant impact on unit economics by increasing the total cost of ownership (TCO) of each reactor (FIG. 1). Moreover, each cycle requires an investment in catalyst, hardware, feed and labor, which is calculated based on the forecast yield. If the yield is down and downtime is up, then return on investment (ROI) will be lower. The impacts of maldistribution are unpredictable, and unpredictability is difficult to budget.

FIG. 1. The presence of hotspots (A) that can reduce distillate (diesel) yield. Localized hotspots (B) can lead to over-cracking and the conversion of desired products (e.g., distillates) into undesirable light ends.

Spotting maldistribution: It is all in the temperature. The key to identifying maldistribution is monitoring and understanding temperature variations in the catalyst bed (FIG. 2A). The development of hotspots is an important indicator that a reactor is suffering from maldistribution.

FIG. 2. Onset of maldistribution shown by increasing and more variable radial ∆Ts (A). Maldistribution may cause localized hotspots to form, resulting in high radial ∆Ts (B).

Hotspots are problematic, as they increase the weighted average bed temperature (WABT) of a catalyst bed. This results in the EOR temperature being reached sooner than expected, and the early termination of the catalyst cycle life.

The pattern (or distribution) of hotspots provides valuable information about the probable cause of maldistribution (FIG. 2B). For example, if the whole north side of a reactor is hotter than the south side, this suggests that the north (hotter) side is experiencing restricted liquid/gas flow, while the south (colder) side experiences higher liquid/gas flow, possibly because of a tilted catalyst bed surface or a tilted internal component, such as a distribution tray.

What is (really) causing your maldistribution?

The first response of most reactor operators whose units are suffering maldistribution is to think that their hardware is to blame. For example, they may assume that their distribution trays have not been installed properly or that they are faulty. However, more often than not, the problem does not lie in the hardware, but somewhere many do not look: in the quality of the catalyst loading.

Catalyst loading controls four key variables that are central to the liquid/gas flow and thermal characteristics of a catalyst bed: the density, homogeneity, profile of the catalyst bed and the catalyst integrity. Get any of these wrong and, no matter how robust the processes are or how well the hardware is performing, the reactor will likely suffer maldistribution and reduced performance.

Even with industry-leading distribution trays that offer near-perfect catalyst wetting (FIG. 3), poor-quality catalyst loading can reduce the performance advantages offered by optimized distribution hardware.

FIG. 3. Well-designed distribution trays can offer near-perfect catalyst wetting, essential for optimal liquid/gas distribution.

Low average catalyst bed density. The average density of the catalyst bed describes how tightly individual catalyst particles are packed together throughout the entire bed (FIG. 4). If the average density of the bed is too low, then the total mass of the loaded catalysts will be lower, resulting in reduced overall catalyst activity.

FIG. 4. Poor-quality catalyst loading results in poorly aligned catalyst particles (left), larger void spaces and a lower total mass of the loaded catalysts. High-quality catalyst loading (right) results in well-aligned catalyst particles, smaller void spaces and a higher total mass of loaded catalysts.

Low homogeneity. Homogeneity describes how the localized density of the catalyst bed varies throughout its volume. Low homogeneity reflects significant localized density variations, even if the average density of the bed meets expectations. If local density varies, then fluids will preferentially flow through lower-density zones, thereby leaving higher-density zones poorly utilized and prone to the development of hotspots. Moreover, WHSV is lower than average in high-density zones, while it is higher than average (not to say significantly higher) in low-density zones.

Poor catalyst integrity. Catalyst integrity describes the extent to which catalyst particles remain intact during loading. Broken catalyst particles form fines and dust that plug the interstitial space between particles, and prevent liquid and gas from freely and homogeneously circulating between all catalyst particles. Increases in local density, caused by fines and dust, can also lead to higher pressure drops across the catalyst bed, which can significantly affect the reactor catalyst cycle length.

Uneven bed profile. If the top surface of the catalyst bed is tilted or uneven, then liquid/gas will preferentially converge toward the lowest point of the top surface and create a preferential flow path (FIG. 5). This will occur even if the overlying distributor trays are well-installed, perfectly leveled and running optimally.

FIG. 5. Sock loading and conventional dense loading often lead to an uneven catalyst bed top surface that channels fluid flow through preferential zones (light blue shading). High-quality homogeneous dense loading creates a flat top surface that, in combination with optimized distributor trays, enables even wetting of the catalyst volume.

What controls the quality of catalyst loading?

The quality of catalyst loading is ultimately controlled by two key factors: operator’s mindset and technology choice.

How an operator’s mindset impacts catalyst loading. When considering the most important factors that determine reactor performance, most operators will list their operational processes, the reactor internals and the catalyst as their star performers. However, the importance of catalyst loading is often overlooked, despite the significant impact it can have on reactor performance.

When it comes to catalyst loading activities, most refiners want to have them executed as fast as possible and at the lowest possible cost. According to the authors, as many as 90% of refinery personnel are uninformed about the signs, causes and impacts of poorly loaded catalyst; this is a conversation that commonly only happens when a reactor is already suffering liquid/gas maldistribution.

How technology choice impacts catalyst loading. The desire for cheap and quick catalyst loading often dictates the choice of loading technology. Two methods are in widespread use: sock loading and dense loading.

Sock loading. Sock loading involves the loading of catalyst via a large canvas tube (FIG. 5) and is the favored method for operators looking for a low-cost loading solution. However, despite its relatively low cost, sock loading has the following significant drawbacks:

  • Low density: With sock loading, the catalyst is essentially “dropped” out of the end of the sock in a manner and at a rate that inhibits the settling of well-aligned catalyst particles. This results in lower-density packing with large void spaces between particles, which reduces the mass of catalyst that can be loaded.
  • Low homogeneity: Improper settling and non-homogeneous distribution of catalyst particles make the catalyst bed vulnerable to differential compaction as particles shift and void spaces collapse under the weight of overlying catalyst. This leads to the formation of localized pockets of higher-density packing and inhomogeneities that promote the development of liquid/gas maldistribution.
  • Uneven profile: Because sock loading only distributes particles over a narrow area, catalyst handlers are required to work inside the reactor to direct the sock. However, despite their best efforts, the result is invariably poor catalyst distribution, a tilted or uneven top catalyst surface, and inhomogeneities caused by the trampling of catalyst underfoot.

Conventional (lower-quality) dense loading. Conventional dense loading (CDL) involves the distribution of catalyst in a rain-like fashion from a particle dispenser lowered through an overlying manway. As the catalyst is allowed to fall freely, the particles settle in a more stable and uniform way, thus increasing bed density and homogeneity. Yet, although CDL technologies—some of which are decades old—are generally superior to sock loading, they also have drawbacks:

  • Poor catalyst integrity and inhomogeneity: CDL technology typically uses a set of rapidly rotating whips to propel catalyst outward, forming a diverging shower of particles. However, the whips are very aggressive, and they reduce catalyst integrity by breaking the catalyst into shorter (finer) particles that settle in denser, random arrangements, resulting in inhomogeneities. This impact is greater in larger reactors where greater forces are needed to propel the catalyst over longer distances to the reactor walls.
  • Uneven profile: The most widely used conventional dense-loading machines often suffer from irregular distribution and suboptimal load rates, and lack the power to propel catalyst to the edges of the reactor—a problem amplified in larger units. This results in uneven mountain, doughnut or valley surface profiles that generate preferential flow channels (FIG. 5). To prevent this, personnel are frequently required to climb inside the reactor to redistribute the catalyst.
  • Space restriction: Whip mechanisms require more vertical and radial space to operate, which makes them difficult to deploy in smaller-diameter reactors. Moreover, they enable catalyst to be loaded only to 70 cm (28 in.) below the overlying distribution tray; a sock is then needed to finish the loading to the required elevation.

Moreover, CDL is unable to compensate for the shadow effects caused by installed hardware, such as thermocouple support structures, vertical quench and catalyst dump pipes. The shadow effect caused by these structures will often lead to maldistribution due to variations in bulk density in the vicinity of the installed hardware.

Homogeneous (high-quality) dense loading. Homogeneous dense loading (HDL) with an advanced catalyst loading methodb operates on the same principles as CDL, but provides a significantly higher quality of loading due to a unique catalyst-distributing mechanism that relies on centrifugal force and brushes to distribute catalyst gently and evenly across a range of reactor sizes: from less than 0.5 m to more than 8 m (19 in. to 26 ft) in diameter (FIG. 6).

FIG. 6. HDL technology (left) is designed to load catalyst at the optimal rate and disperse particles, using centrifugal force gently and evenly across the full diameter of the catalyst bed (right).

HDL can also minimize the shadow effects of internals when used with optimally designed hardware. For example, optimized internals can minimize the shadow effects of the hardware installed within the catalyst bed, which helps minimize the risk of liquid/gas maldistribution. Using leaner hardware with large and easy-to-open manways also facilitates higher-quality and faster loading.

HDL has many advantages over sock and CDL methods, including:

  • Improved density: By using an optimal loading rate up to 30 m3/hr (1,050 ft3/hr), HDL can achieve bed densities up to 30% higher vs. sock loading and CDL, and can thus load as much as 30% more catalyst by weight (FIG. 7).
    FIG. 7. Catalyst loading rate (A) has an important influence on the average bed density achieved. Higher average density means that a great total mass of catalyst (B) can be loaded within the same reactor space.  
  • Increased homogeneity: Optimal load rates enable catalyst particles to fall freely and separately across the full diameter of any reactor. In doing so, particles have time to settle in more stable, consistent and horizontal arrangements, resulting in more homogeneous loading density over the entire bed volume. This improves the liquid/gas distribution and lowers the risk of hotspot development within the catalyst bed. Additionally, HDL can load to within 30 cm of the overlying dispersion trays, thereby eliminating the need for top-up by sock loading. This reduces the risk of liquid/gas maldistribution and hotspot development.
  • Excellent catalyst integrity: HDL uses low-friction materials and brushes at the mouths of the dispersion channels to minimize damage to catalyst particles, thereby eliminating nearly all dust and fine particles.

Strategic benefits of HDL

HDL provides multiple strategic benefits that make it the smart choice for operators looking to maximize unit and hardware performance, and ROI.

Improved safety. Safety should be the top priority for any operator. By eliminating the need for personnel to enter the reactor and by only requiring a single person to install and dismantle the machinery, HDL enables a much safer environment for catalyst handlers.

Higher yields. Enhanced liquid/gas flow distribution, optimized WHSV and improved vapor-liquid-catalyst interactions maximize product output per unit of run time.

Lower TCO. Although the initial price may be higher than alternative solutions, HDL can generate a healthy ROI by providing a cost-effective way to safeguard the performance of expensive catalyst and reactor internals, and to also reduce lifetime maintenance costs.

Longer run time and less downtime. With no maldistribution, EOR temperatures are not reached prematurely, so the best use is made of the catalyst activity (FIG. 8). Additionally, the efficiency with which HDL machines can be loaded and dismantled can save as much as two shifts’ worth of labor and time over the course of a cycle. Case Study 2 provides an example of how unit run time can be extended by using HDL.

FIG. 8. Example of EOR times for the three loading methods: sock loading, CDL and HDL.

Reassurance. Knowing that catalyst is well loaded means one less thing for the operator to worry about.

How to minimize the risk of maldistribution

To obtain the best performance out of expensive catalyst and state-of-the-art reactor internals, an increasing number of operators are unlocking major business value by embracing the long-term strategic benefits of high-quality HDL. To achieve this, reactor operators should:

  • Be proactive in understanding the importance of catalyst loading in reactor performance—getting catalyst loading right at the start can save a lot of time, money and stress in the long term.
  • Speak with experts on catalyst loading and reactor internals to better understand how to maximize the performance of fixed-bed catalytic reactors.
  • Think long term to ensure that short-term goals do not compromise the overall performance and success of the project.

Case Study 1: Eliminating hotspots

A process engineer for a mineral oil plant in Brazil reported that significant temperature anomalies (hotspots) had developed in a small, 0.9-m-diameter reactor. Using data from pairs of temperature sensors positioned inside the catalyst bed, the engineer calculated a temperature increase of as much as 20°C (68°F) from one side of the reactor to the other—a clear indication of maldistribution.

Deciding that the catalyst needed reloading, the engineer opted for the HDL method and technology. Following reloading, the temperature variation across the reactor was reduced to 3°C (37°F), which indicated that the maldistribution had been eliminated—a highly satisfactory result for the engineer (FIG. 9).

FIG. 9. A reactor records severe temperature anomalies down its left side after sock loading or CDL (left). After reloading catalyst using HDL, temperature anomalies are almost eliminated (right).  

Case Study 2: Extending run time

A reactor operator at an oil refinery in the Czech Republic was planning to reload catalyst for two gasoil hydrotreater reactors of 2.8 m and 3.3 m diameters. Previously, loading was done using whip-based CDL; however, because of restrictive internal configurations, the quality of the loading was suboptimal.

The operator instead chose the HDL method and technology, and, after 7 mos, observed better performance in terms of reduced WABT. Based on new projections, the time to reach the EOR temperature of 395°C (743°F) was extended significantly beyond the initial 14-mos target (FIG. 10). HP

FIG. 10. Graph showing the evolution of WABT and extension of EOR time.


a Shell’s high-dispersion (HD) tray
b CREALYST-Oil’s CALYDENS® advanced catalyst loading system

The Authors

Related Articles

From the Archive



{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}